Positioning anomaly detection method, system and apparatus for automated guided vehicle, positioning compensation method, system and apparatus for automated guided vehicle, and storage medium

Abstract

Provided are a positioning anomaly detection method, system and apparatus for an automated guided vehicle, a positioning compensation method, system and apparatus for an automated guided vehicle, and a storage medium. The positioning anomaly detection method comprises: acquiring a first pose deviation of a first automated guided vehicle relative to each of a plurality of first docking objects; acquiring a second pose deviation of each of one or more second automated guided vehicles relative to each of a plurality of second docking objects; on the basis of the first pose deviations and the second pose deviations, determining a pose deviation similarity between the first automated guided vehicle and each second automated guided vehicle; and on the basis of the pose deviation similarities, determining whether a positioning apparatus of the first automated guided vehicle is abnormal. The positioning compensation method comprises: acquiring a reference first positioning error of a target automated guided vehicle docking with a target docking position, wherein the reference first positioning error is caused by a positioning apparatus of the target automated guided vehicle; acquiring a reference second positioning error of the target automated guided vehicle docking with the target docking position, wherein the reference second positioning error is caused by the calibration of the target docking position; on the basis of the reference first positioning error and the reference second positioning error, compensating for the target docking position, so as to acquire a compensated target docking position; and on the basis of the compensated target docking position, controlling the automated guided vehicle to move towards the compensated target docking position.

Description

Positioning anomaly detection and compensation method, system, device and storage medium of automated mobile equipment Cross-reference to related applications

[0001] This application claims priority to Chinese application No. 202411268737.1 filed on September 10, 2024, and Chinese application No. 202511093453.8 filed on August 5, 2025, the entire contents of which are incorporated herein by reference. TECHNICAL FIELD

[0002] The present specification relates to the technical field of automated mobile equipment, in particular to a positioning anomaly detection and compensation method, system, device and storage medium of automated mobile equipment. BACKGROUND

[0003] Before the automated mobile equipment (AGV) docks with the docking object, the AGV needs to be positioned by a positioning device and plan a route to the vicinity of the docking object, so as to dock with the docking object. However, the installation position of the positioning device on the automated mobile equipment may slowly shift over time, resulting in errors in the calibration results of the positioning device, and thus the positioning, navigation and other information obtained by the positioning device also inevitably have errors. If the error is large, the automated mobile equipment cannot operate normally, and even the overall operation state of the docking site may be destroyed.

[0004] Currently, periodic manual calibration or fixed threshold alarm is usually used to suppress the deviation. However, the operation and maintenance personnel need to pause production regularly, and calibrate each AGV by external instruments, which is high in labor cost and affects production capacity; and the fixed threshold strategy cannot distinguish between "normal running noise" and "device slow drift", and is prone to false alarms or missed alarms when the scene changes (such as new rack layout, ground reflection change).

[0005] Therefore, it is necessary to provide a positioning anomaly detection and compensation method, system, device and storage medium of automated mobile equipment, which mines the pose deviation data of multiple vehicles and common docking objects, adaptively calculates the pose deviation similarity, and then judges whether the automated mobile equipment is abnormal, so as to realize accurate, real-time and low-cost monitoring of the positioning device anomaly. SUMMARY

[0006] One or more embodiments of the present specification provide a positioning anomaly detection method of an automatic mobile device, the method comprising: obtaining a first pose deviation of a first automatic mobile device relative to each first docking object in a plurality of first docking objects; obtaining a second pose deviation of each second automatic mobile device in one or more second automatic mobile devices relative to each second docking object in a plurality of second docking objects; determining a pose deviation similarity between the first automatic mobile device and the each second automatic mobile device based on the first pose deviation and the second pose deviation; and determining whether a positioning device of the first automatic mobile device has an anomaly based on the pose deviation similarity.

[0007] One or more embodiments of the present specification provide a positioning anomaly detection system of an automatic mobile device, the system comprising: an obtaining module configured to obtain a first pose deviation of a first automatic mobile device relative to each first docking object in a plurality of first docking objects; the obtaining module is further configured to obtain a second pose deviation of each second automatic mobile device in one or more second automatic mobile devices relative to each second docking object in a plurality of second docking objects; a determining module configured to determine a pose deviation similarity between the first automatic mobile device and the each second automatic mobile device based on the first pose deviation and the second pose deviation; and a judging module configured to determine whether a positioning device of the first automatic mobile device has an anomaly based on the pose deviation similarity.

[0008] One or more embodiments of the present specification provide a positioning anomaly detection device of an automatic mobile device, the device comprising at least one processor and at least one memory; the at least one memory is configured to store computer instructions; the at least one processor is configured to execute at least part of the computer instructions to implement a positioning anomaly detection method of an automatic mobile device.

[0009] One or more embodiments of the present specification provide a computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, a positioning anomaly detection method of an automatic mobile device is implemented.

[0010] One or more embodiments of the present specification provide a positioning compensation system of an automatic mobile device, the system comprising: an error obtaining module configured to obtain a reference first positioning error of a target automatic mobile device docking a target docking position, the reference first positioning error caused by a positioning device of the target automatic mobile device; the error obtaining module is further configured to obtain a reference second positioning error of the target automatic mobile device docking the target docking position, the reference second positioning error caused by a calibration of the target docking position; a compensation module configured to compensate the target docking position based on the reference first positioning error and the reference second positioning error to obtain a compensated target docking position; and a control module configured to control the automatic mobile device to move to the compensated target docking position based on the compensated target docking position.

[0011] One or more embodiments of the present specification provide a positioning compensation device of an automatic mobile device, the device comprising at least one processor and at least one memory; the at least one memory is configured to store computer instructions; the at least one processor is configured to execute at least part of the computer instructions to implement a positioning compensation method of an automatic mobile device.

[0012] One or more embodiments of the present specification provide a computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, a positioning compensation method of an automatic mobile device is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0013] The present specification will be further described in the manner of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, the same numbers represent the same structures, wherein:

[0014] FIG. 1 is a schematic diagram of an application scenario of a positioning anomaly detection system of an automatic mobile device according to some embodiments of the present specification;

[0015] FIG. 2 is a schematic diagram of exemplary hardware and software components of an exemplary computing device according to some embodiments of the present specification;

[0016] FIG. 3 is an exemplary module diagram of a positioning anomaly detection system of an automatic mobile device according to some embodiments of the present specification;

[0017] FIG. 4 is an exemplary flowchart of a positioning anomaly detection method of an automatic mobile device according to some embodiments of the present specification;

[0018] FIG. 5 is an exemplary schematic diagram of determining a pose deviation similarity according to some embodiments of the present specification;

[0019] FIG. 6 is an exemplary schematic diagram for determining pose deviation similarity based on clustering, according to some embodiments of the present specification;

[0020] FIG. 7 is an exemplary schematic diagram for determining pose deviation similarity based on candidate pose deviation similarity, according to some embodiments of the present specification;

[0021] FIG. 8 is an exemplary flow chart of a method for detecting positioning anomaly of an automated mobile device, according to another embodiment of the present specification;

[0022] FIG. 9 is an exemplary block diagram of a positioning compensation system of an automated mobile device, according to some embodiments of the present specification;

[0023] FIG. 10 is an exemplary flow chart of a positioning compensation method of an automated mobile device, according to some embodiments of the present specification;

[0024] FIG. 11 is another exemplary flow chart of a positioning compensation method of an automated mobile device, according to some embodiments of the present specification;

[0025] FIG. 12 is another exemplary flow chart of a positioning compensation method of an automated mobile device, according to some embodiments of the present specification. DETAILED DESCRIPTION

[0026] The schemes of the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0027] In the following description, specific details such as specific system structures, interfaces, techniques, etc. are presented in order to provide a thorough understanding of the present application for the sake of explanation, rather than for the sake of limitation.

[0028] The term "and / or" herein is merely an association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects. In addition, "multiple" herein means two or more than two. In addition, the term "at least one" herein means any combination of any one or more of the multiple, or any combination of at least two of the multiple, for example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0029] In the docking process of an automated mobile device (e.g., AGV, AMR, etc.), a positioning device completes path planning and navigation by collecting environmental data in real time, so that the guided vehicle can be accurately parked near the target docking object.

[0030] However, due to slow displacement or calibration error accumulation of the positioning device in long-term use, there is a non-negligible deviation between the actual parking pose and the ideal pose, which affects the docking accuracy and even interferes with the overall operation rhythm on site. Relying on manual periodic calibration or fixed threshold alarm not only consumes manpower, but also is difficult to adapt to complex and variable operation scenarios.

[0031] Therefore, some embodiments of the present specification provide a positioning anomaly detection method, system and device medium of an automatic mobile device, which can flexibly calculate the positioning similarity based on multi-vehicle historical docking data, adaptively identify whether the positioning device is abnormal, significantly improve the detection accuracy and scene adaptability, and ensure the long-term stable operation of the automatic mobile device.

[0032] FIG. 1 is a schematic diagram of an application scenario of a positioning anomaly detection system of an automatic mobile device according to some embodiments of the present specification.

[0033] In some embodiments, as shown in FIG. 1, the application scenario 100 of the positioning anomaly detection system of the automatic mobile device can include an automatic mobile device 110, a docking object 120, a network 130, a storage device 140, and a processor 150. In some embodiments, the positioning anomaly detection system of the automatic mobile device can be implemented by the methods and / or processes disclosed in the present specification.

[0034] The automatic mobile device 110 can be a robotic device that completes tasks such as material handling, picking, docking, etc. by autonomous movement with little or no human involvement. The automatic mobile device 110 can be applied to production lines, warehouse logistics, pharmaceutical delivery, commercial services, etc. For example, an Automated Guided Vehicle (AGV), an Autonomous Mobile Robot (AMR), etc.

[0035] In some embodiments, the automatic mobile device 110 is configured to physically interact with the docking object 120 to complete a specific operation or task.

[0036] The docking object 120 can be an object that the automatic mobile device needs to interact with to perform a task or complete a specific operation. The task or operation can include material handling, picking, sorting, data transmission, material loading, unloading, assembly, charging, parking, door opening, elevator access, etc. For example, the docking object 120 can be a shelf, a handle, a socket, an elevator, an automatic door, a robotic arm, or other devices that require the automatic mobile device to provide or take away materials, and the embodiments of the present specification do not limit this. The automatic mobile device 110 can perform the corresponding task after completing docking with the docking object 120.

[0037] In some embodiments, the autonomous mobile device 110 can implement movement and navigation based on waypoints. Waypoints can be used to define paths, trigger actions of the autonomous mobile device 110, etc. For example, a waypoint can be set with certain attributes, e.g., which can represent an action that the autonomous mobile device 110 performs at the waypoint, e.g., park, type of operation performed at the waypoint (e.g., pick, load, unload, charge, etc.), associated action (e.g., which signal to trigger after reaching the waypoint), etc. Waypoints can include task waypoints, path waypoints, wait waypoints, charge waypoints, etc. A task waypoint is a waypoint at which the autonomous mobile device 110 performs a task. A path waypoint is a waypoint used by the autonomous mobile device 110 to plan or optimize a path.

[0038] In some embodiments, the docking object 120 can include waypoints. For example, the autonomous mobile device 110 can dock at a docking waypoint.

[0039] In some embodiments, for example, for AMR, a waypoint is a specific location coordinate recorded and named on a reference map. In some embodiments, for example, for AGV, a waypoint is a pre-set location with physical markers on a fixed path. The physical markers can be used to guide the AGV to move and / or perform corresponding actions. The fixed path can be guided by magnetic strips or magnetic nails. The physical markers can include RFID tags, optical tags, magnetic nail sequences, etc. The RFID tags can be buried at regular intervals or key locations on the fixed path. When the RFID reader on the AGV passes the tag, the unique ID number stored in the RFID tag will be read. This ID number corresponds to a waypoint. The central control system has pre-defined what each ID number represents (i.e. attributes or positions, such as stopping points, right turns, left turns, etc.). The optical tags can include color bands, color patterns, QR codes, etc. After the camera on the AGV recognizes the optical tags, it can determine its own location. The magnetic nail sequence can include a specific number or arrangement of magnetic nails, which determines the attributes or positions corresponding to the waypoint (e.g. "three consecutive magnetic nails represent a charging point"). In some embodiments, the automated mobile device 110 can achieve the recognition and / or positioning of the waypoint based on positioning technology. The positioning technology can include guided positioning technology and / or autonomous positioning technology. The guided positioning technology can include using physical markers to achieve positioning. For example, the guided positioning technology can include magnetic guidance technology, visual recognition technology, retro-reflective board laser navigation technology, laser contour matching technology, etc. The retro-reflective board laser navigation technology can scan the pre-set retro-reflective board around the waypoint by laser, and measure the position and orientation relative to the retro-reflective board based on the reflected signal of the retro-reflective board. The visual recognition technology can use a camera to recognize physical markers such as AprilTag, ArUco markers or specific visual features on the docking object to obtain images, and calculate the relative position and angular deviation from the physical markers by analyzing the physical markers in the images. The laser contour matching technology can use 2D / 3D laser radar to scan the contour of the docking object (such as the shape of the charging pile), and match it with the pre-set model to find the accurate relative pose. The autonomous positioning technology can include SLAM navigation technology. For example, laser SLAM navigation technology, visual SLAM navigation technology, etc. The SLAM navigation technology (simultaneous localization and mapping) can construct a map (local map) of the surrounding environment according to sensor data such as laser radar (LiDAR), vision sensor (camera), inertial measurement unit (IMU), etc., and estimate the position in the pre-set reference map (i.e. global map) according to the local map.

[0040] In some embodiments, the automated mobile device can be provided with a positioning device for navigation.

[0041] The positioning device can include a laser radar, a vision sensor, etc.

[0042] In some embodiments, in response to the positioning device determining the position information of the docking object through the navigation technology, the processor can control an automated guided vehicle (AGV) to go to the docking object according to a fixed track.

[0043] In some embodiments, in response to the positioning device determining the position information of the docking object through the navigation technology, the processor can calculate a reasonable path from the current position of the autonomous mobile robot (AMR) to the position of the docking object, and control the autonomous mobile robot to move to the docking object according to the reasonable path.

[0044] The network 130 can include any suitable network that can facilitate the exchange of information and / or data. In some embodiments, one or more components of the application scenario 100 can exchange information and / or data through the network 130. For example, the processor 150 can obtain point cloud data scanned by the automatic mobile device 110 through the network 130. The network 130 can include a local area network (LAN), a wide area network (WAN), a wired network, a wireless network, etc., or any combination thereof.

[0045] The storage device 140 can store data or information generated by other devices. In some embodiments, the storage device 140 can store data and / or information scanned by the automatic mobile device 110, such as point cloud data, etc. In some embodiments, the storage device 140 can store data and / or information processed by the processor 150, such as the first pose deviation, etc. The storage device 140 can include one or more storage components, each of which can be a separate device or a part of other devices. The storage device can be local or implemented through the cloud. In some embodiments, the storage device 140 can include a mass storage, a removable storage, a volatile read-write memory, a read-only memory (ROM), etc., or any combination thereof. In some embodiments, the storage device 140 can be implemented on a cloud platform. In some embodiments, the storage device 140 can be part of the processor 150.

[0046] The processor 150 can be configured to manage data resources and process data and / or information from various modules of the application scenario 100 and / or external data sources of the positioning anomaly detection system of the autonomous mobile device. The processor 150 can execute program instructions based on the data, information and / or processing results, thereby performing one or more functions described in the present specification. For example, the processor 150 can obtain a first pose deviation of a first autonomous mobile device relative to each first docking object in a plurality of first docking objects; obtain a second pose deviation of each second autonomous mobile device in one or more second autonomous mobile devices relative to each second docking object in a plurality of second docking objects; determine a pose deviation similarity between the first autonomous mobile device and each second autonomous mobile device based on the first pose deviation and the second pose deviation; and determine whether the positioning device of the first autonomous mobile device is abnormal based on the pose deviation similarity.

[0047] In some embodiments, the processor 150 can be a single server or a group of servers. The group of servers can be centralized or distributed. In some embodiments, the processor 150 can be local or remote. In some embodiments, the processor 150 can be implemented on a cloud platform. By way of example only, the cloud platform can include a private cloud, a public cloud, a hybrid cloud, or any combination thereof.

[0048] FIG. 2 is a schematic diagram of exemplary hardware and software components of an exemplary computing device, according to some embodiments of the present specification.

[0049] The computing device 200 can be a general-purpose computer or a special-purpose computer, both of which can be used to implement the positioning anomaly detection method of the autonomous mobile device and the positioning compensation method of the autonomous mobile device described in some embodiments of the present specification. The computing device 200 can include various modules for implementing the positioning anomaly detection system of the autonomous mobile device and the positioning compensation system of the autonomous mobile device as described in the present specification. For example, the processor can be implemented on the computing device 200 by its hardware, software program, firmware or combination thereof. For the sake of convenience, only one computer is shown in the figure, but the computer functions related to the positioning anomaly detection method of the autonomous mobile device and the positioning compensation method of the autonomous mobile device described in the present specification can be implemented in a distributed manner on multiple similar platforms to distribute the processing load.

[0050] For example, the computing device 200 can include a communication port 250 that is connected to and / or from a network to enable data communication. The computing device 200 can also include a processor 220 in the form of one or more processors to execute program instructions. An example computer platform can include an internal communication bus 210, different types of memory storage and data storage (e.g., a disk 270, a read only memory (ROM) 230, or a random access memory (RAM) 240), various data files that are processed and / or transmitted by the computer. An example computer platform also includes program instructions stored in the ROM 230, the RAM 240, and / or other forms of non-transitory storage media that are executed by the processor 220. The methods and / or processes of the present specification can be implemented in the form of program instructions. The computing device 200 can also include an input / output (I / O) interface 260 that can support input / output between the computer and other components. The computing device 200 can also receive programming and data through network communication. More about the network can be found in the description related to FIG. 1.

[0051] For illustration only, only one CPU and / or processor is exemplarily described in the computing device 200. However, it should be noted that the computing device 200 in some embodiments of the present specification can include multiple CPUs and / or processors, and thus the operations and / or methods described in the present specification as implemented by one CPU and / or processor can also be implemented by multiple CPUs and / or processors jointly or independently. For example, if in the present specification, the CPU and / or processor of the computing device 200 performs operation A and operation B, it should be understood that operation A and operation B can also be performed by two different CPUs and / or processors in the computing device 200 jointly or independently (e.g., a first processor performs operation A, a second processor performs operation B, or the first and second processors jointly perform operation A and operation B).

[0052] Those skilled in the art can understand that the structure shown in FIG. 2 is only a block diagram of part of the structure related to the schemes of the present specification, and does not constitute a limitation on the computer device to which the schemes of the present specification are applied. A specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0053] FIG. 3 is a module diagram of a positioning anomaly detection system of an autonomous mobile device according to some embodiments of the present specification.

[0054] In some embodiments, as shown in FIG. 3, the positioning anomaly detection system 300 of the autonomous mobile device can include an acquisition module 310, a determination module 320, and a judgment module 330.

[0055] In some embodiments, the obtaining module 310 is configured to obtain a first pose deviation of the first automatic mobile device relative to each first docking object in the plurality of first docking objects; and obtain a second pose deviation of each second automatic mobile device in the one or more second automatic mobile devices relative to each second docking object in the plurality of second docking objects.

[0056] In some embodiments, the determining module 320 is configured to determine, based on the first pose deviations of at least part of the corresponding first docking objects in the plurality of first docking objects and the second pose deviations of at least part of the corresponding second docking objects in the plurality of second docking objects, a pose deviation similarity between the first automatic mobile device and each second automatic mobile device.

[0057] In some embodiments, the determining module 320 is further configured to determine a total number of same docking objects in the plurality of first docking objects and the plurality of second docking objects; determine, based on the total number, a target first pose deviation of the first automatic mobile device and a target second pose deviation of each second automatic mobile device for determining the pose deviation similarity, the target first pose deviation comprising a first pose deviation of the first automatic mobile device relative to a target first docking object in the plurality of first docking objects, and the target second pose deviation comprising a second pose deviation of each second automatic mobile device relative to a target second docking object in the plurality of second docking objects; and determine, based on the target first pose deviation and the target second pose deviation, the pose deviation similarity between the first automatic mobile device and each second automatic mobile device.

[0058] In some embodiments, the determining module 320 is further configured to, in response to determining that the total number is less than a first threshold, determine the first pose deviations of the plurality of first docking objects as the target first pose deviation, and determine the second pose deviations of the plurality of second docking objects as the target second pose deviation; in response to determining that the total number is greater than the first threshold, determine the first pose deviations of at least part of the same docking objects as the target first pose deviation, and determine the second pose deviations of at least part of the same docking objects as the target second pose deviation.

[0059] In some embodiments, the determining module 320 is further configured to, in response to determining that the total number is greater than the first threshold and less than a second threshold, determine the first pose deviations of the same docking objects as the target first pose deviation, and determine the second pose deviations of the same docking objects as the target second pose deviation; and in response to determining that the total number is greater than the second threshold, determine the first pose deviations of part of the same docking objects as the target first pose deviation, and determine the second pose deviations of part of the same docking objects as the target second pose deviation.

[0060] In some embodiments, the determining module 320 is further configured to determine a first cluster center by clustering the target first pose deviations; determine a second cluster center by clustering the target second pose deviations; and determine the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first cluster center and the second cluster center.

[0061] In some embodiments, the determining module 320 is further configured to determine a candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the at least part of the same docking objects; and determine the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the candidate pose deviation similarity corresponding to the at least part of the same docking objects.

[0062] In some embodiments, the determining module 320 is further configured to determine a candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the same docking objects; and determine the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the candidate pose deviation similarity corresponding to the same docking objects.

[0063] In some embodiments, the determining module 320 is further configured to determine a plurality of candidate pose deviation similarities based on the first pose deviation and the second pose deviation; and determine the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the plurality of candidate pose deviation similarities.

[0064] In some embodiments, the determining module 320 is further configured to determine a first representative pose deviation based on the first pose deviation; determine a second representative pose deviation based on the second pose deviation; and determine the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first representative pose deviation and the second representative pose deviation.

[0065] In some embodiments, the determining module 320 is further configured to determine a target pose similarity based on the pose deviation similarities between the first automated mobile device and the plurality of second automated mobile devices; and determine whether the positioning apparatus of the first automated mobile device is abnormal based on the target pose similarity.

[0066] In some embodiments, the judging module 330 is configured to determine whether the positioning apparatus of the first automated mobile device is abnormal based on the pose deviation similarity.

[0067] For more information about the above-mentioned modules, please refer to FIG. 4-FIG. 8 and the related descriptions.

[0068] It should be noted that the above description of the positioning anomaly detection system 300 of the automatic mobile device and its modules is for the convenience of description only, and cannot limit the scope of the present specification to the embodiments described. It can be understood that, for those skilled in the art, after understanding the principles of the device, any combination of the modules or connection of the sub-devices with other modules can be made without departing from the principles. In some embodiments, the acquisition module 310, the determination module 320 and the judgment module 330 in FIG. 3 can be different modules in one device, or one module can implement the functions of two or more modules described above. For example, the modules can share one storage module, or each module can have its own storage module. Variations such as these are within the scope of protection of the present specification.

[0069] FIG. 4 is an exemplary flowchart of a positioning anomaly detection method of an automatic mobile device according to some embodiments of the present specification. In some embodiments, the flow 400 is executed by a processor. The flow 400 includes steps 410-440.

[0070] Step 410, acquiring a first pose deviation of a first automatic mobile device relative to each first docking object in a plurality of first docking objects.

[0071] The first automatic mobile device refers to an automatic mobile device for which it is necessary to determine whether there is a positioning anomaly. For example, the first automatic mobile device can be an automatic guided vehicle (AGV).

[0072] The first docking object refers to an object to which the first automatic mobile device is docked. For example, the first docking object can be a shelf, a handle, a socket, etc. to which the first automatic mobile device is docked, which is not limited here.

[0073] The first pose deviation refers to the pose deviation of the first automatic mobile device relative to the first docking object. The first pose deviation refers to the deviation between the pose perceived by the first automatic mobile device after reaching the target point corresponding to the first docking object and the ideal pose corresponding to the target point.

[0074] In some embodiments, the first pose deviation can include a position deviation and / or an attitude deviation. The position deviation refers to the difference between the position of the automatic mobile device after movement and the position of the docking object. The attitude deviation can refer to the included angle value between the orientation of the automatic mobile device after movement and the orientation of the docking object.

[0075] In some embodiments, the processor can determine the first pose deviation in various ways.

[0076] For example, when the first automatic mobile device reaches the target point (target docking position) corresponding to the first docking object, the processor can scan the first docking object (e.g., the positioning code of the first docking object) through an imaging device (e.g., a laser camera, a laser radar) to obtain point cloud data, and determine the position and pose of the first automatic mobile device relative to the first docking object at the target point based on the point cloud data, i.e., the first pose deviation data.

[0077] When the first automatic mobile device reaches the target point, the processor can emit laser to the first docking object through a laser radar, and obtain an image (i.e., point cloud data) of the first docking object formed by the laser through an imaging device (e.g., a CMOS sensor); the processor can calculate the bending or displacement of the laser line in the field of view of the imaging device caused by the height change of the surface of the first docking object based on the point cloud data, and calculate the three-dimensional coordinates of each point on the surface of the first docking object based on the bending or displacement in the field of view of the imaging device caused by the point cloud data, so as to obtain the 2D or 3D contour of the first docking object, and determine the position and pose of the first automatic mobile device relative to the first docking object based on the 2D or 3D contour of the first docking object and the position and orientation of the 2D or 3D contour of the first docking object in the point cloud data, i.e., the first pose deviation.

[0078] For another example, when the first automatic mobile device reaches the target point, the processor can scan the positioning code of the first docking object through a scanning device (e.g., an image code reader, a laser code scanner, a depth camera, an RFID reader, etc.), and accurately know the orientation and position of the first automatic mobile device according to the coordinate information contained in the positioning code and the deflection angle of the first automatic mobile device relative to the positioning code, so as to determine the first pose deviation of the first automatic mobile device relative to the first docking object.

[0079] The positioning code refers to a component that provides accurate position information to the automatic mobile device. For example, the positioning code can include a two-dimensional code, a magnetic stripe, a magnetic pin, an RFID tag, a barcode, an AprilTag, etc. A unique ID or coordinate information is pre-stored in each positioning code.

[0080] In some embodiments, the scanning device and the imaging device can be arranged on the automatic mobile device (e.g., in the first automatic mobile device and the second automatic mobile device).

[0081] In some embodiments, for each first docking object, the processor can first obtain a first historical pose deviation of the first automatic mobile device from each first docking object at a plurality of first historical time points, determine a first pose deviation based on the first historical pose deviations at the plurality of first historical time points, for example, can determine a central tendency representation value corresponding to the plurality of first historical pose deviations as the first pose deviation. The central tendency representation value can be one of the median, mean, mode, variance, etc. of the first historical pose deviations, which is not specifically limited here. The first historical time point refers to a historical time point before the current time point corresponding to the first automatic mobile device.

[0082] In some embodiments, each first automatic mobile device can save, package the first historical pose deviation of the first docking object at each first historical time point, and upload the packaged file to a local storage or a cloud storage. When obtaining the first historical pose deviation, the processor can query the storage file corresponding to the first automatic mobile device according to the first docking object and the first automatic mobile device in the local storage or the cloud storage, and then query the historical pose deviation corresponding to the first docking object in the storage file as the first historical pose deviation. When the first historical pose deviations are statistically summed, the sum value is divided by the total number of the first historical pose deviations to obtain the central tendency representation value (i.e. mean) corresponding to the first historical pose deviations, which is taken as the first pose deviation. The first docking object and the first automatic mobile device can be a number or other unique data representation, which is not specifically limited here.

[0083] Step 420, obtaining a second pose deviation of each second automatic mobile device from each second docking object in a plurality of second docking objects.

[0084] The second automatic mobile device refers to other automatic mobile devices in the same transportation scenario as the first automatic mobile device. In some embodiments, the second automatic mobile device can be used to determine whether the positioning of the first automatic mobile device is abnormal. The second automatic mobile device has the same configuration as the first automatic mobile device. For example, the first automatic mobile device and the second automatic mobile device use the same positioning navigation technology and the same positioning device.

[0085] The second docking object refers to the object docked by the second automatic mobile device. For example, the second docking object can be a shelf, an operating handle, a socket, etc. docked by the second automatic mobile device, which is not specifically limited here.

[0086] The second pose deviation refers to a pose deviation of the second automatic mobile device relative to the second docking object. The second pose deviation refers to a deviation between a perceived pose of the second automatic mobile device after the second automatic mobile device reaches a target point corresponding to the second docking object and an ideal pose corresponding to the target point.

[0087] In some embodiments, the second pose deviation can include a position deviation and an attitude deviation. More details about the position deviation and the attitude deviation can be found in the foregoing relevant description.

[0088] In some embodiments, the second pose deviation is obtained in the same way as the first pose deviation. More details about obtaining the first pose deviation can be found in the foregoing relevant description.

[0089] In some embodiments, the processor can determine the second pose deviation in various ways.

[0090] For example, when the second automatic mobile device reaches the target point corresponding to the second docking object, the processor can scan the second docking object through the imaging device to obtain point cloud data, and determine the position and attitude of the second automatic mobile device at the target point relative to the second docking object based on the point cloud data, i.e., the second pose deviation data.

[0091] When the second automatic mobile device reaches the target point, the processor can emit laser to the second docking object through the laser radar, and obtain an image (i.e., point cloud data) of the laser formed on the second docking object through the imaging device; the processor can calculate a bending or displacement of the laser line in the field of view of the imaging device caused by a height change of the surface of the second docking object based on the point cloud data, and calculate three-dimensional coordinates of each point on the surface of the second docking object based on the bending or displacement in the field of view of the imaging device caused by the point cloud data, so as to obtain a 2D or 3D contour of the second docking object, and calculate a position and an orientation of the 2D or 3D contour of the second docking object in the point cloud data, so as to determine the position and the attitude of the second automatic mobile device relative to the second docking object, i.e., the second pose deviation.

[0092] For another example, when the second automatic mobile device reaches the target point, the processor can scan a positioning code of the second docking object through the scanning device, and accurately know the orientation and the position of the second automatic mobile device according to coordinate information contained in the positioning code and a deflection angle of the second automatic mobile device relative to the positioning code, so as to determine the second pose deviation of the second automatic mobile device relative to the second docking object.

[0093] In some embodiments, for each second docking object, the processor can first obtain a second historical pose deviation of the second automatic mobile device from the second docking object at each of a plurality of second historical time points, determine a second pose deviation based on the second historical pose deviations at the plurality of second historical time points, for example, can determine a central tendency representation value corresponding to the plurality of second historical pose deviations as the second pose deviation. The central tendency representation value can be one of the median, mean, mode, variance, etc. of the second historical pose deviations, which is not specifically limited herein. The second historical time point refers to a historical time point before the current time point corresponding to the second automatic mobile device.

[0094] In some embodiments, each second automatic mobile device can save, package the second historical pose deviation of the second docking object at each second historical time point, and upload the packaged file to a local storage or a cloud storage. When obtaining the second historical pose deviation, the processor can query the storage file corresponding to the second automatic mobile device according to the second docking object and the second automatic mobile device in the local storage or the cloud storage, and then query the historical pose deviation corresponding to the second docking object in the storage file as the second historical pose deviation. When the plurality of second historical pose deviations are statistically summed, the sum is divided by the total number of the second historical pose deviations to obtain the central tendency representation value (i.e. mean) corresponding to the second historical pose deviations, which is taken as the second pose deviation. The second docking object and the second automatic mobile device can be a number or other unique data representation, which is not specifically limited herein.

[0095] The processor can obtain the first historical pose deviation and the second historical pose deviation in various ways.

[0096] For example, the processor can first obtain a reference docking pose (also referred to as an ideal docking pose) when the current automatic mobile device moves to a target point according to a planned path. Wherein the current automatic mobile device is the first automatic mobile device or the second automatic mobile device, and the planned path is generated based on the collected data of the positioning device on the current automatic mobile device (for example, the target point corresponding to the positioning code can be obtained based on the camera scanning the positioning code of the docking object (the first docking object or the second docking object), and the path to the target point is determined based on the target point). The target point represents the position of the current automatic mobile device when docking with the docking object. When the automatic mobile device moves to the target point, the scanning device (for example, a laser radar) on the automatic mobile device scans the positioning code or the docking object to obtain point cloud data, and the actual docking pose of the current automatic mobile device relative to the docking object can be determined based on the point cloud data. Finally, the pose deviation between the actual docking pose and the reference docking pose is obtained, and the pose deviation is taken as the historical pose deviation (the first historical pose deviation or the second historical pose deviation). The actual docking pose refers to the pose of the automatic mobile device perceived by the automatic mobile device relative to the docking object. The reference docking pose refers to the pose that the automatic mobile device needs to reach when docking with the target docking object. The target point and the reference docking pose can be system default settings, and each docking object can correspond to a reference docking pose. The target point and the reference docking pose corresponding to the docking object can be obtained by scanning the positioning code.

[0097] For example, when the automatic mobile device enters the docking distance range of the docking object, the processor can scan the surrounding environment and the docking object by using the positioning device (such as a laser radar) to obtain point cloud data, the automatic mobile device generates the center coordinates of the docking object according to the point cloud data, and plans a path from the current position of the automatic mobile device to the docking object according to the center coordinates of the docking object, and tracks the planned path. When the current automatic mobile device approaches the position of the docking object (such as the edge of the docking object), the image acquisition device on the automatic mobile device is started to acquire the image of the positioning code of the docking object, and the target point and the reference docking pose corresponding to the docking object are determined according to the acquired image. Further, the automatic mobile device can modify the end point of the planned path based on the target point, and re-plan the path and track the new planned path based on the modified end point. When the automatic mobile device reaches the target point, the camera scans the positioning code again to obtain the actual docking pose of the automatic mobile device. Finally, the actual docking pose and the reference docking pose are compared to obtain the pose deviation, and the pose deviation is taken as the first historical pose deviation or the second historical pose deviation.

[0098] For example, when the automatic mobile device moves to the target point according to the planned path, the processor can first capture the docking object by using a scanning device (such as a camera) to obtain a docking image; then analyze the docking image, extract the positioning code features of the docking object, and calculate the center coordinates and the reference docking pose; then, the automatic mobile device calculates the pose deviation according to the reference docking pose and the current pose of the automatic mobile device, and corrects the end point of the planned path according to the pose deviation; then, the automatic mobile device re-plans the path based on the corrected end point and tracks the new path until it reaches the target point again; at this time, the scanning device is triggered again to scan the positioning code to obtain the actual docking pose of the automatic mobile device; finally, the actual docking pose and the reference docking pose are compared to obtain the pose deviation, and the pose deviation is taken as the first historical pose deviation or the second historical pose deviation.

[0099] For example, when the automatic mobile device moves to the target point according to the planned path, the image acquisition device is used to capture the docking object to obtain a docking image. Then, the automatic mobile device analyzes the docking image to obtain the center coordinates of the docking object, and compares the coordinate position of the automatic mobile device at this time with the center coordinates of the docking object to obtain the pose deviation between the current automatic mobile device and the docking object, and take the pose deviation as the first historical pose deviation or the second historical pose deviation.

[0100] In step 430, based on the first pose deviation corresponding to at least part of the plurality of first docking objects and the second pose deviation corresponding to at least part of the plurality of second docking objects, the pose deviation similarity between the first automatic mobile device and each second automatic mobile device is determined.

[0101] The pose deviation similarity refers to the similarity between the first pose deviation and the second pose deviation.

[0102] After obtaining the first pose deviation and the second pose deviation, the pose deviation similarity between the first automatic mobile device and the second automatic mobile device can be obtained in various ways according to the first pose deviation and the second pose deviation. For example, the pose deviation similarity between the first pose deviation and the second pose deviation can be calculated by using Pearson correlation coefficient, Euclidean distance, cosine similarity, Manhattan distance, etc. It can be understood that the calculation method of the pose deviation similarity can also use methods such as Jaccard similarity coefficient and Chebyshev distance, which are not limited here.

[0103] For more details, please refer to the related description of FIGS. 5-8.

[0104] In some embodiments, the processor can determine a plurality of candidate pose deviation similarities based on the first pose deviation and the second pose deviation; and determine the pose deviation similarity between the first mobile device and each second mobile device based on the plurality of candidate pose deviation similarities.

[0105] In some embodiments, the processor can determine a plurality of candidate pose deviation similarities based on the first pose deviation and the second pose deviation; and determine the pose deviation similarity between the first mobile device and each second mobile device based on the plurality of candidate pose deviation similarities.

[0106] In some embodiments, the processor can determine a centralized trend representation value (e.g., mean, mode, etc.) of the plurality of candidate pose deviation similarities as the pose deviation similarity.

[0107] In some embodiments of the present disclosure, the pose deviation similarity between the first mobile device and each second mobile device is determined based on the first pose deviation and the second pose deviation, which can make full use of historical data of all docking objects, avoid interference of individual abnormal samples on the result, and improve the robustness and accuracy of the similarity calculation in the statistical level, thereby reducing the false negative rate and false positive rate of the positioning device.

[0108] In some embodiments, the processor can determine a first representative pose deviation based on the first pose deviation of the first mobile device relative to the plurality of first docking positions; determine a second representative pose deviation of each second mobile device based on the second pose deviation of each second mobile device relative to the plurality of second docking positions; and determine the pose deviation similarity between the first mobile device and each second mobile device based on the first representative pose deviation and the second representative pose deviation.

[0109] The first representative pose deviation refers to a pose deviation obtained by processing the plurality of first pose deviations. For example, the first representative pose deviation can be a mean, variance, maximum value, minimum value, median, mode, etc. of the plurality of first pose deviations.

[0110] The second representative pose deviation refers to a pose deviation obtained by processing the plurality of second pose deviations of the second mobile device relative to the plurality of second docking positions. For example, the second representative pose deviation can be a mean, variance, maximum value, minimum value, median, mode, etc. of the plurality of second pose deviations.

[0111] In some embodiments, the processor can determine the first representative pose deviation based on the plurality of first pose deviations relative to the plurality of first docking positions by statistical calculation. For example, the processor can determine the mean or the mode of the plurality of first pose deviations as the first representative pose deviation based on the plurality of first pose deviations by statistical calculation. The second representative pose deviation is determined in the same way as the first representative pose deviation.

[0112] In some embodiments, the processor can cluster the plurality of first pose deviations to obtain a first cluster center, and determine the first cluster center as the first representative pose deviation. The processor can further cluster the plurality of second pose deviations to obtain a second cluster center, and determine the second cluster center as the second representative pose deviation corresponding to the second mobile device.

[0113] In some embodiments, the processor can perform similarity calculation on the first representative pose deviation and the second representative pose deviation, and determine the calculated similarity as the pose deviation similarity.

[0114] For example, after obtaining each first pose deviation and each second pose deviation, the processor can calculate the variance of each first pose deviation to obtain the first representative pose deviation of the first mobile device, and calculate the variance of each second pose deviation to obtain the second representative pose deviation corresponding to the second mobile device. Then, the processor can calculate the similarity between the first representative pose deviation and the second representative pose deviation, i.e., the pose deviation similarity between the first mobile device and the second mobile device, by using a similarity calculation method such as Euclidean distance, cosine similarity, etc.

[0115] In some embodiments, the processor can determine the total number of the same docking objects in the plurality of first docking objects and the plurality of second docking objects, determine the target first pose deviation of the first mobile device and the target second pose deviation of each second mobile device for determining the pose deviation similarity based on the total number, and determine the pose deviation similarity between the first mobile device and each second mobile device based on the target first pose deviation and the target second pose deviation. For more information about determining the pose deviation similarity between the first mobile device and each second mobile device, please refer to FIG. 5 and the related description thereof.

[0116] In some embodiments of the present disclosure, the first representative pose deviation is determined based on the first pose deviation, the second representative pose deviation is determined based on the second pose deviation, and the pose deviation similarity is determined based on the first representative pose deviation and the second representative pose deviation, which can greatly reduce the data dimension and the calculation amount, improve the real-time performance, effectively filter out incidental noise, highlight the system drift trend, and make the abnormality determination more focused and more accurate, while taking into account the detection efficiency and reliability.

[0117] Step 440, based on the pose deviation similarity, determining whether the positioning device of the first automatic mobile device is abnormal.

[0118] In some embodiments, when the number of the second automatic mobile devices is one, the processor can compare the pose deviation similarity between the first automatic mobile device and the second automatic mobile device with a preset normal value. If the pose deviation similarity is equal to the preset normal value, it indicates that the positioning device of the first automatic mobile device is normal. Otherwise, it indicates that the positioning device of the first automatic mobile device is abnormal. The preset normal value can be a system default setting or a manual setting.

[0119] In some embodiments, when the number of the second automatic mobile devices is multiple, the processor can determine a target pose similarity based on multiple pose deviation similarities between the first automatic mobile device and the multiple second automatic mobile devices; and determine whether the positioning device of the first automatic mobile device is abnormal based on the target pose similarity.

[0120] The processor can determine the target pose similarity based on the multiple pose deviation similarities between the first automatic mobile device and the multiple second automatic mobile devices in multiple ways.

[0121] For example, the processor can take the mean value of the multiple pose deviation similarities between the first automatic mobile device and the multiple second automatic mobile devices as the target pose similarity.

[0122] For another example, the processor can take the maximum value or the mode value of the multiple pose deviation similarities between the first automatic mobile device and the multiple second automatic mobile devices as the target pose similarity.

[0123] The processor can determine whether the target pose similarity exceeds a similarity threshold value based on the target pose similarity and the similarity threshold value. If not, it is determined that the positioning device is abnormal. The similarity threshold value is a parameter for determining whether the positioning device is abnormal. In some embodiments, the similarity threshold value can be preset by a technician based on experience.

[0124] In some embodiments, the processor can compare the plurality of pose deviation similarities between the first automatic mobile device and the plurality of second automatic mobile devices with a similarity threshold, and count the number of pose deviation similarities greater than the similarity threshold. When the proportion of pose deviation similarities greater than the similarity threshold (the ratio of the number of pose deviation similarities greater than the similarity threshold to the total number of pose deviation similarities) is greater than a proportion threshold, it can be determined that the positioning device is normal; when the proportion of pose deviation similarities greater than the similarity threshold (the ratio of the number of pose deviation similarities greater than the similarity threshold to the total number of pose deviation similarities) is less than the proportion threshold, it can be determined that the positioning device is abnormal.

[0125] In some embodiments of the present specification, based on the pose deviation similarities between the first automatic mobile device and the plurality of second automatic mobile devices, the target pose similarity is determined; and then it is determined whether the positioning device of the first automatic mobile device is abnormal, which can adapt to different scenarios and different batches of equipment differences, significantly reducing the cost of manual calibration and threshold parameter adjustment; at the same time, using historical pose deviation similarities for statistical decision can detect small drifts in the early stage and avoid the risk of docking failure or production stoppage caused by error accumulation, thereby improving the consistency and long-term operation reliability of the automatic mobile device.

[0126] In some embodiments, the processor can perform similarity calculation on the first automatic mobile device and the plurality of second automatic mobile devices to obtain a plurality of pose deviation similarities, and then determine whether the positioning device of the first automatic mobile device is abnormal based on the plurality of pose deviation similarities. For example, the processor can filter a plurality of pose deviation similarities satisfying a similarity requirement from the pose deviation similarities corresponding to the plurality of second automatic mobile devices; count the plurality of pose deviation similarities to obtain a representative pose deviation similarity between the first automatic mobile device and the plurality of second automatic mobile devices; and determine whether the positioning device of the first automatic mobile device is abnormal based on the representative pose deviation similarity.

[0127] In some embodiments, the similarity requirement is that the pose deviation similarity is greater than a first similarity threshold. Specifically, the filtering can be performed using a logical function, and the formula of the logical function is represented by formula (1):

[0128] Wherein, T is the first similarity threshold, and t is the pose deviation similarity; when the pose deviation similarity is greater than or equal to the first similarity threshold, the logical function H(t, T) outputs 1; when the pose deviation similarity is less than the first similarity threshold, the logical function H(t, T) outputs 0.

[0129] For the several pose deviation similarities obtained after screening, the sum of several pose deviation similarities is obtained, and the ratio of the sum of pose deviation similarities to the number of second automatic mobile devices is used as the representative pose deviation similarity.

[0130] In some embodiments, a second similarity threshold is used to determine the similarity of the representative pose deviation to ascertain whether the positioning device of the first automated mobile device is abnormal. If the similarity of the representative pose deviation is less than the second similarity threshold, it is determined that the positioning device of the first automated mobile device is abnormal. The second similarity threshold can be a value such as 0.3, 0.5, or p, and is not specifically limited here. In some embodiments, the first and second similarity thresholds can be preset by a technician based on experience.

[0131] In a specific implementation scenario, the positioning device is a LiDAR, the image acquisition device is a camera, and the docking object is a shelf. Furthermore, the docking process of the automated mobile device can be transformed into digital space, specifically for the device numbered a. j The automated mobile device issues a docking number of s k When performing a docking task on the shelf, the docking task is saved to the first input space X, and then processed by the corresponding first output space Y to obtain the docking pose deviation. The first input space is X = {(a1,s1),...,(a1,s...}. N ),...,(a j ,s k ),...(a M ,s1),...,(a M ,s N The first output space is M represents the number of automated mobile devices, N represents the number of shelves, and 1 ≤ j ≤ M, 1 ≤ k ≤ N. Input x = (a j ,s k )∈X represents the number a j Automatic mobile device docking number s k The task of docking with the shelf is defined by the output y∈Y, which represents the pose deviation between the actual docking pose and the reference docking pose of the automated mobile device. The transition from the first input space to the first output space can be represented as y=f(x).

[0132] With number a j The docking number of the automated mobile equipment is s k Taking the shelf as an example, the specific process for collecting docking task data is as follows: First, at a certain historical moment, numbered a j The automated mobile device received the docking number s k After the task of the shelf is completed, the number is a j Automated mobile devices utilize their onboard lidar to scan the surrounding environment and the area designated s.k The shelf was scanned to obtain point cloud data, numbered a. j Automatic mobile devices generate number s based on point cloud data k The center coordinates of the shelf, and according to the number s k The center coordinates of the shelf are used to plan the layout from the number a j The current location of the automated mobile device to the number s k The planned path of the shelf, and the result of numbering a j The reference docking pose of the automated moving part when it moves to the target point according to the planned path, and tracks the planned path. In the part numbered a j When the automated mobile device is about to reach the target point on the planned path, activate the function with number a. j The camera mounted on the automated mobile device, for the number s k Images are captured from the shelves, and the target points of the planned path are modified based on the captured images. The path is then replanned and tracked to obtain the result numbered a. j The automated mobile device reaches the actual docking pose of the new planned path target point. Finally, the function y obtains the pose deviation bias based on the actual docking pose and the reference docking pose. sample Among them, the pose deviation is... sample As f((a) j ,s k The historical pose deviation of f((a)). To ensure the stability of the model output, information about f((a) will be stored. j ,s k Several historical pose deviations were obtained. Then, the deviation numbered a was obtained. j Automatic mobile equipment with number s k The average value of the shelf, as the pose deviation, is calculated using the following formula: Where L is f((a) j ,s k The number of historical pose deviations.

[0133] The above method obtains the pose deviation of each automated mobile device. When obtaining the pose deviation similarity between different automated mobile devices, this process can also be transformed into a digital space, that is, using the second input space C and the second output space Z to obtain the pose deviation similarity between different automated mobile devices. The second input space is C = {(a1, a1), ..., (a1, a...} M ),...,(a j ,a k ),...(a M ,a1),...,(a M ,a M}, the second output space is Z = [0, 1] ∈ R, where 1≤j≤M, 1≤k≤M, the input 1≤k≤M represents the number a j The automatic mobile device and the automatic mobile device numbered a k form a pair of combinations, and the output z ∈ Z represents the positioning similarity between the automatic mobile devices. From the second input space to the second output space, it can be represented as z = g(c). Where g((a j ,a k )) = g((a k ,a j )), g(a j ,a j ) = 1.

[0134] In order to solve the positioning similarity between the first automatic mobile device numbered a j and the second automatic mobile device numbered a k , the processor can filter out the number of shelves that the first automatic mobile device numbered a j and the second automatic mobile device numbered a k have docked, that is, the number of samples of shelves and , where i = 1,...,n, n is the total number of shelves that the first automatic mobile device numbered a j and the second automatic mobile device numbered a k have docked. According to the size of the total number of target shelves n, different processing is carried out, where the first threshold N1 and the second threshold N2 are respectively the minimum number of shelves and the maximum number of shelves set by humans:

[0135] If n < N1, all shelves that the first automatic mobile device has docked are obtained as target docking shelves, and the first pose deviation of each target docking shelf is obtained. Then, the first pose deviation is clustered to obtain the first clustering center Similarly, all shelves that the second automatic mobile device has docked are obtained as target docking shelves, and the second pose deviation of each target docking shelf is obtained. Then, the second pose deviation is clustered to obtain the second clustering center Then, the cosine similarity between the first clustering center and the second clustering center is obtained, and the candidate pose deviation similarity is obtained as the pose deviation similarity, where the calculation formula of the cosine similarity can be represented by formula (2) obtained from formula (3):

[0136] If N1≤n≤N2, n shelves are taken as target docking shelves, and the first automatic mobile device docking shelf​ The first pose deviation The same operation yields the second pose deviation corresponding to the second automated mobile device. Seek and The cosine similarity between the targets is used to obtain the candidate pose deviation similarity. Then, the average of the candidate pose deviation similarities corresponding to the docking racks for each target is calculated to obtain the pose deviation similarity.

[0137] If n > N2, then randomly select N2 shelves from the n shelves as target shelves, and then perform similarity calculation.

[0138] Then, a certain frequency (e.g., half a day or a day) can be set to periodically calculate the similarity of pose deviations between each automated mobile device, and then anomaly detection can be performed on the lasers of M automated mobile devices respectively. Let the detection number be a... j Taking the case of whether the laser of an automated mobile device is abnormal, we obtain the number a. j The similarity between an automated mobile device and other automated mobile devices (excluding itself) is obtained as [g((a j ,a1)),...,g((a j ,a i ),...,g((a j ,a M ))], where i = 1,...,M, and i ≠ j. Define logical functions. And set the first similarity threshold Thre. The number a... j The similarity between an automated mobile device and other automated mobile devices (excluding itself) is substituted into the following function: Where i ≠ j. The second similarity threshold is set to 0.5. If result < 0.5, then the number is determined to be a. j The lidar of the automated mobile device is malfunctioning; if result ≥ 0.5, then the determination number is a. j The lidar of the automated mobile device is functioning normally.

[0139] In some embodiments of this specification, by obtaining the first pose deviation between the first automated mobile device and the first docking object, and obtaining the second pose deviation between the second automated mobile device and the second docking object, and then using each first pose deviation and each second pose deviation to calculate the displacement deviation similarity between the first automated mobile device and the second automated mobile device, the similarity of the pose deviation can be determined. By using the pose deviation similarity, it can be determined whether there is an abnormality in the positioning device of the first automated mobile device, thereby accurately detecting whether there is an abnormality in the positioning device.

[0140] It should be noted that the above description of flow 400 is merely for example and illustration, and does not limit the scope of the present specification. Various modifications and changes can be made to flow 400 by those skilled in the art under the guidance of the present specification. However, these modifications and changes are still within the scope of the present specification.

[0141] FIG. 5 is an exemplary flowchart of determining pose deviation similarity, according to some embodiments of the present specification. In some embodiments, flow 500 can be executed by a processor. Flow 500 includes steps 510-530.

[0142] Step 510, determine the total number of same docking objects in the plurality of first docking objects and the plurality of second docking objects.

[0143] In some embodiments, the first docking object and the second docking object can be the same docking object. The embodiments of the present specification mainly apply to the field of automatic mobile devices, by calculating the pose deviation similarity between the first pose deviation of the same docking object which the first automatic mobile device has historically docked with, and the second pose deviation of the same docking object which the second automatic mobile device has historically docked with, to confirm whether the positioning device installed on the first automatic mobile device is abnormal. The first automatic mobile device or the second automatic mobile device can perform operations such as carrying and data transmission on the docking object after completing docking with the docking object.

[0144] In some embodiments, the first docking object and the second docking object can include the same docking object and include different docking objects, i.e., there is an intersection between the first docking object and the second docking object. Further, in the case of multiple second automatic mobile devices, the number of docking objects corresponding to different second automatic mobile devices can be the same or different. The pose deviation similarity between the first automatic mobile device and different second automatic mobile devices is also not the same.

[0145] For example, the first automatic mobile device has docked with docking objects numbered 1, 2, 3, 4, and 5, the second automatic mobile device has docked with docking objects numbered 2, 3, 4, 5, and 6, and the third automatic mobile device has docked with docking objects numbered 1, 4, 5, and 7. Therefore, for the first automatic mobile device and the second automatic mobile device, the same docking objects are docking objects numbered 2, 3, 4, and 5; for the first automatic mobile device and the third automatic mobile device, the same docking objects are docking objects numbered 1, 4, and 5; and for the second automatic mobile device and the third automatic mobile device, the same docking objects are docking objects numbered 4 and 5.

[0146] At step 520, based on the total number of the same docking objects in the plurality of first docking objects and the plurality of second docking objects, a target first pose deviation of the first automatic mobile device for determining the pose deviation similarity and a target second pose deviation of each second automatic mobile device are determined.

[0147] The target first pose deviation refers to the first pose deviation for determining the pose deviation similarity. In some embodiments, the target first pose deviation can include at least one. The target first pose deviation can be the first pose deviation selected from the plurality of first pose deviations.

[0148] The target second pose deviation refers to the second pose deviation for determining the pose deviation similarity. In some embodiments, the target second pose deviation can include at least one. The target second pose deviation can be the second pose deviation selected from the plurality of second pose deviations.

[0149] In some embodiments, in response to determining that the total number is less than the first threshold, the processor can determine the first pose deviation of the plurality of first docking objects as the target first pose deviation, and determine the second pose deviation of the plurality of second docking objects as the target second pose deviation; in response to determining that the total number is greater than or equal to the first threshold, determine the first pose deviation of at least part of the same docking objects as the target first pose deviation, and determine the second pose deviation of at least part of the same docking objects as the target second pose deviation.

[0150] In some embodiments, in response to determining that the total number is greater than or equal to the first threshold, the processor can determine the first pose deviation of at least part of the same docking objects as the target first pose deviation based on a first preset number, and determine the second pose deviation of at least part of the same docking objects as the target second pose deviation based on the first preset number. The first preset number can be the same as the total number of the same docking objects, or can be less than the total number of the same docking objects.

[0151] The first threshold and the first preset number can be a system default value, an empirical value, a human pre-set value, or any combination thereof, and can be set according to actual needs, which is not limited in the present specification.

[0152] In some embodiments of the present specification, the target first pose deviation and the target second pose deviation are determined based on the relationship between the total number and the first threshold, which avoids false positives in small sample scenarios, and improves real-time performance in large-scale scenarios, thereby significantly optimizing overall resource efficiency while ensuring accuracy of anomaly detection.

[0153] In some embodiments, in response to determining that the total number is greater than the first threshold value and less than the second threshold value, the processor can determine the first pose deviation of the same docking object as the target first pose deviation, and determine the second pose deviation of the same docking object as the target second pose deviation; and in response to determining that the total number is greater than or equal to the second threshold value, determine the first pose deviation of part of the same docking object as the target first pose deviation, and determine the second pose deviation of part of the same docking object as the target second pose deviation.

[0154] In some embodiments, in response to determining that the total number is greater than or equal to the second threshold value, the processor can determine the first pose deviation of part of the same docking object as the target first pose deviation based on the second preset number, and determine the second pose deviation of part of the same docking object as the target second pose deviation based on the second preset number. The second preset number is less than the total number of the same docking object.

[0155] The second threshold value and the second preset number can be a system default value, an empirical value, a human pre-set value, or any combination thereof, and can be set according to actual needs, which are not limited in the present specification. In some embodiments, the second preset number can be equal to the first preset number.

[0156] In some embodiments, the number of part of the same docking object (i.e., the second preset number) can be equal to the second threshold value.

[0157] It can be understood that, in response to determining that the total number of the same docking object is greater than or equal to the second threshold value, the processor determines the first pose deviation of the same docking object of the second threshold value as the target first pose deviation, and determines the second pose deviation of the same docking object of the second threshold value as the target second pose deviation.

[0158] In some embodiments of the present specification, the target first pose deviation and the target second pose deviation are determined based on the total number, the first threshold value and the second threshold value, which can ensure statistical sufficiency; when the data amount further increases, the sampling mode is automatically switched, which significantly reduces the calculation, storage and communication overhead, while still maintaining sufficient representation. Avoiding the risk of underfitting in a small sample scenario, and preventing resource waste in a large sample scenario, thereby achieving a dynamic balance between detection accuracy and system efficiency, and improving the long-term scalability of automatic mobile device anomaly monitoring.

[0159] Step 530, determining the pose deviation similarity between the first automatic mobile device and each second automatic mobile device based on the target first pose deviation and the target second pose deviation.

[0160] In some embodiments, the processor can determine the pose deviation similarity based on similarity calculation between the target first pose deviation and the target second pose deviation.

[0161] If the total number of the same docking objects in the plurality of first docking objects and the plurality of second docking objects is less than the first threshold, the processor can take all the first pose deviations corresponding to the plurality of first docking objects as the target first pose deviations, and take all the second pose deviations corresponding to the plurality of second docking objects as the target second pose deviations. If the total number of the same docking objects in the plurality of first docking objects and the plurality of second docking objects is greater than the first threshold and less than or equal to the second threshold, the processor can determine the first pose deviations and the second pose deviations corresponding to the same docking objects in the plurality of first docking objects and the plurality of second docking objects as the target first pose deviations and the target second pose deviations. If the total number of the same docking objects in the plurality of first docking objects and the plurality of second docking objects is greater than the second threshold, the processor can take the central tendency value (such as the mean, median, maximum, etc.) of the first pose deviations corresponding to the same docking objects in the plurality of first docking objects and the plurality of second docking objects as the target first pose deviations, and take the central tendency value (such as the mean, median, maximum, etc.) of the second pose deviations corresponding to the same docking objects as the target second pose deviations.

[0162] In some embodiments of the present specification, by first counting the intersection number of the two batches of docking objects, and then adaptively selecting the target pose deviations according to the intersection size, both the statistical reliability and the calculation efficiency can be considered, that is, the full samples can be fully utilized when the data is scarce, and the sampling rate can be automatically reduced when the data is redundant. Then, the pose deviation similarity is calculated based on the refined target deviations, which can significantly reduce noise interference and computational complexity.

[0163] In some embodiments, the processor can cluster the target first pose deviations to determine a first clustering center of the target first pose deviations, cluster the target second pose deviations to determine a second clustering center corresponding to the target second pose deviations, and determine the pose deviation similarity between the first automatic mobile device and each second automatic mobile device based on the first clustering center and the second clustering center. In some embodiments, the processor can obtain the total number of the same docking objects in the first docking objects and the second docking objects; if the total number is less than the first threshold, the pose deviation similarity between the first automatic mobile device and each second automatic mobile device is determined based on the first clustering center and the second clustering center.

[0164] In some embodiments, the processor can determine the candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the at least part of the same docking objects, and determine the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity corresponding to the at least part of the same docking objects. In some embodiments, the processor can obtain the total number of the same docking objects in the first docking objects and the second docking objects, and determine the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity corresponding to the at least part of the same docking objects if the total number of the same docking objects is greater than a first threshold.

[0165] In some embodiments, the processor can determine the candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the same docking objects, and determine the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity corresponding to the same docking objects. In some embodiments, the processor can obtain the total number of the same docking objects in the first docking objects and the second docking objects, and determine the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity corresponding to the same docking objects if the total number of the same docking objects is greater than a first threshold and less than a second threshold.

[0166] The first threshold is less than the second threshold, and the processor can select the second threshold number of target docking objects from the plurality of target docking objects as the screened target docking objects if the total number is greater than the second threshold.

[0167] For example, the total number of the same docking objects docked by the first automatic mobile device and the second automatic mobile device is n, the first threshold is N1, and the second threshold is N2. N1 and N2 are default values, and N1 < N2. If n < N1, all docking objects (i.e., the first docking objects) docked by the first automatic mobile device are obtained, the first pose deviation of each docking object is obtained, and then the first pose deviation is clustered to obtain the first clustering center center1. Similarly, all docking objects (i.e., the second docking objects) docked by the second automatic mobile device are obtained, the second pose deviation of each representative docking object is obtained, and then the second pose deviation is clustered to obtain the second clustering center center2. Then, the cosine similarity between the first clustering center center1 and the second clustering center center2 (i.e., the pose deviation similarity between the first automatic mobile device and the second automatic mobile device) is calculated, which can be represented by the following formula (3):

[0168] If N1≤n≤N2, all same docking objects to which the first and second automatic mobile devices dock are obtained, and each first pose deviation and each second pose deviation corresponding to the same docking object are obtained, and then the first pose deviation and the second pose deviation are one-to-one corresponding calculated for the candidate pose deviation similarity according to the same docking object, and the pose deviation similarity between the first automatic mobile device and the second automatic mobile device is determined based on the candidate pose deviation similarity.

[0169] For example, the first pose deviation and the second pose deviation corresponding to the same docking object numbered 1 are corresponding, and the candidate pose deviation similarity between the first pose deviation and the second pose deviation corresponding to the docking object numbered 1 is calculated; the first pose deviation and the second pose deviation corresponding to the same docking object numbered 2 are corresponding, and the candidate pose deviation similarity between the first pose deviation and the second pose deviation corresponding to the docking object numbered 2 is calculated. Then, the average value of the candidate pose deviation similarity corresponding to each docking object is calculated, which is the central tendency representation value, and the average value is taken as the pose deviation similarity.

[0170] If n>N2, the second threshold number of docking objects in the docking objects are selected, that is, N2 docking objects are randomly selected, and then each first pose deviation and each second pose deviation corresponding to the docking objects are obtained, and the first pose deviation and the second pose deviation are corresponding according to the docking objects, and the candidate pose deviation similarity is calculated, and then the average value of the candidate pose deviation similarity corresponding to each docking object is calculated, which is the central tendency representation value, and the average value is taken as the pose deviation similarity.

[0171] FIG. 6 is an exemplary schematic diagram for determining the pose deviation similarity based on clustering, according to some embodiments of the present specification.

[0172] In some embodiments, as shown in FIG. 6, the processor can cluster the target first pose deviation 540 to determine a first cluster center 610, cluster the target second pose deviation 550 to determine a second cluster center 620, and determine the pose deviation similarity 560 between the first automatic mobile device and each second automatic mobile device based on the first cluster center 610 and the second cluster center 620.

[0173] The first cluster center refers to the target first pose deviation that is in the cluster center after clustering the target first pose deviation.

[0174] The second cluster center refers to the target second pose deviation that is in the cluster center after clustering the target second pose deviation.

[0175] In some embodiments, the processor can cluster the target first pose deviation and the target second pose deviation in various ways.

[0176] For example, the processor can perform clustering analysis on the target first pose deviation and the target second pose deviation respectively by a clustering algorithm (such as K-means, DBSCAN, hierarchical clustering, etc.), to obtain a first clustering center and a second clustering center.

[0177] In some embodiments, the processor can determine the pose deviation similarity based on the first clustering center and the second clustering center in multiple ways.

[0178] For example, after obtaining each target first pose deviation (for example, determining the first pose deviation of the first automatic mobile device to all the first docking objects that have been docked as the target first pose deviation) and each target second pose deviation (for example, determining the second pose deviation of the second automatic mobile device to all the second docking objects that have been docked as the target first pose deviation), the processor can cluster each first pose deviation to obtain a first clustering center, and take the first clustering center as a first representative pose deviation of the first automatic mobile device, and then cluster each second pose deviation to obtain a second clustering center, and take the second clustering center as a second representative pose deviation corresponding to the second automatic mobile device. Then, the similarity between the first representative pose deviation and the second representative pose deviation is calculated to obtain the pose deviation similarity.

[0179] In some embodiments of the present specification, determining the pose deviation similarity based on the first clustering center and the second clustering center can significantly compress the data amount and reduce the computational complexity; at the same time, the clustering center has statistical robustness, which can effectively suppress the influence of abnormal samples and measurement noise, making the subsequent calculation of the pose deviation similarity more stable and accurate, thereby improving the reliability and real-time performance of the positioning device abnormality detection.

[0180] In some embodiments, the total number of the same docking objects can be less than a first threshold.

[0181] It can be understood that if the total number of the same docking objects of the first target mobile device and the second target mobile device is less than the first threshold, the processor can determine the first clustering center and the second clustering center, and determine the pose deviation similarity between the first automatic mobile device and each second automatic mobile device based on the first clustering center and the second clustering center.

[0182] In some embodiments of the present specification, in the case of a small total number, determining the pose deviation similarity through clustering analysis can avoid the risk of underfitting in a small sample scenario.

[0183] FIG. 7 is an exemplary schematic diagram showing determination of the pose deviation similarity based on the candidate pose deviation similarity according to some embodiments of the present specification.

[0184] In some embodiments, as shown in FIG. 7, the processor can determine a candidate pose deviation similarity 710 based on the target first pose deviation 540 and the target second pose deviation 550 of each of the at least part of the same docking objects; and determine the pose deviation similarity 560 between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity 710 corresponding to the same docking objects.

[0185] The candidate pose deviation similarity refers to a one-to-one corresponding deviation similarity between the target first pose deviation and the target second pose deviation.

[0186] In some embodiments, the processor can determine the similarity between each target first pose deviation and each target second pose deviation as the candidate pose deviation similarity.

[0187] In some embodiments, the processor can take a central tendency value (such as a mean value) of the plurality of candidate pose deviation similarities as the pose deviation similarity.

[0188] In some embodiments of the present specification, determining the pose deviation similarity based on the candidate pose deviation similarity can significantly reduce the delay and resource occupation, thereby improving the accuracy of anomaly detection while taking into account the real-time performance and scalability of large-scale scenarios.

[0189] In some embodiments, the total number of the same docking objects of the first automatic mobile device and the second automatic mobile device is greater than a first threshold value.

[0190] It can be understood that in response to the total number of the same docking objects being greater than the first threshold value, the processor can determine the pose deviation similarity based on the candidate pose deviation similarity.

[0191] In some embodiments of the present specification, when the number of the same docking objects of the first automatic mobile device and the second automatic mobile device exceeds the first threshold value, the system can have sufficient statistical samples to directly aggregate (such as average, mode, median) without clustering, simplifying the algorithm process, reducing the computing power consumption, and ensuring the statistical significance and robustness of the similarity estimation, thereby further improving the real-time performance and resource efficiency while ensuring the detection accuracy.

[0192] In some embodiments, the processor can determine a candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the same docking objects; and determine the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the candidate pose deviation similarity corresponding to the same docking objects.

[0193] In some embodiments, the processor can determine the candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the same docking objects.

[0194] In some embodiments of the present disclosure, by independently calculating the candidate pose deviation similarity for each of the same docking objects and then statistically aggregating all the candidate values, all available data is fully utilized and information loss caused by sampling or dimension reduction is avoided.

[0195] In some embodiments, the total number of the same docking objects is greater than a first threshold and less than a second threshold.

[0196] It can be understood that, in response to the total number of the same docking objects being greater than the first threshold and less than the second threshold, the processor can determine the candidate pose deviation similarity based on the target first pose deviation and the target second pose deviation of each of the same docking objects.

[0197] In some embodiments of the present disclosure, when the number of the same docking objects is between the first threshold and the second threshold, the amount of data is sufficient to support statistical significance and is not too large. At this time, the processor directly utilizes all the same docking objects to generate the candidate pose deviation similarity one by one and aggregates, which avoids information loss caused by sampling and saves additional calculations such as clustering.

[0198] FIG. 8 is an exemplary flowchart of an anomaly detection method according to another embodiment of the present disclosure.

[0199] In some embodiments, as shown in FIG. 8, the flow 800 can be executed by the processor. The flow 800 includes steps 810-850.

[0200] Step 810, based on the first automatic mobile device, traversing the storage space to screen other automatic mobile devices meeting the comparison condition as the second automatic mobile device.

[0201] In some embodiments, the comparison condition can be set as at least one docking object that other automatic mobile devices in the storage space have ever docked with is a docking object that the first automatic mobile device has ever docked with. For example, the first automatic mobile device has docked with docking objects numbered 1, 2, 3, and 4, and then a search is performed in the storage space to find that an automatic mobile device numbered P has docked with docking objects numbered 1, 3, and 5, and an automatic mobile device numbered Q has docked with docking objects numbered 5, 6, and 7. The docking objects that the first automatic mobile device has docked with are compared with the docking objects that the automatic mobile devices numbered P and Q have docked with respectively, and it is determined that the docking objects that the automatic mobile device numbered Q has docked with do not include the docking objects that the first automatic mobile device has docked with, while the automatic mobile device numbered P has docked with docking objects numbered 1 and 3, which are the same as the docking objects that the first automatic mobile device has docked with. Therefore, the automatic mobile device numbered P can be used as the second automatic mobile device.

[0202] In this embodiment, other automatic mobile devices that have no intersection with the docking objects that the first automatic mobile device has ever docked with are regarded as interference items and are excluded, so as to reduce the calculation amount.

[0203] In step 820, a first historical pose deviation of the first automatic mobile device with respect to each target docking object and a second historical pose deviation of the second automatic mobile device with respect to each target docking object are obtained. The first automatic mobile device and the second automatic mobile device are each provided with a positioning device for navigation.

[0204] In some embodiments, after the second automatic mobile device is determined, the first historical pose deviation of the first automatic mobile device with respect to each target docking object and the second historical pose deviation of the second automatic mobile device with respect to each target docking object can be obtained in the storage space according to the label of the first automatic mobile device and the label of the second automatic mobile device.

[0205] In step 830, an average value of each first historical pose deviation is calculated to obtain a first pose deviation, and an average value of each second historical pose deviation is calculated to obtain a second pose deviation.

[0206] In some embodiments, since the first automatic mobile device can have docked with a same target docking object multiple times, in order to reduce the calculation amount, an average value of multiple first historical pose deviations of the first automatic mobile device with respect to the same target docking object can be calculated to obtain a first pose deviation of the first automatic mobile device with respect to the target docking object. Similarly, first pose deviations of the first automatic mobile device with respect to other target docking objects can also be obtained. Similarly, first pose deviations of the second automatic mobile device with respect to each target docking object can also be obtained.

[0207] At step 840, a pose deviation similarity between the first mobile device and the second mobile device is determined based on the first pose deviation and the second pose deviation.

[0208] At step 850, whether the positioning device of the first mobile device is abnormal is determined based on the pose deviation similarity.

[0209] In some embodiments of the present disclosure, by evaluating the similarity between the positioning devices through the docking pose deviation data, the positioning device with low similarity to other positioning devices is determined to be abnormal, and the positioning of the mobile device is fed back to the on-site maintenance personnel, which helps to timely troubleshoot problems on site. Moreover, the pose deviation similarity is calculated based on the docking pose deviation data. The pose deviation similarity is part of the consistency of multiple vehicles, and the similarity is applied to each business process, which helps to improve the consistency of multiple vehicles of the mobile device in each business process.

[0210] Those skilled in the art can understand that, in the above method of the specific implementation, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0211] The mobile device (for example, a robot, etc.) is a comprehensive system integrating environment perception, dynamic decision planning, behavior control and execution, etc., with the advantages of high efficiency, safety and flexibility, etc., and provides a reliable solution for cargo handling. By introducing the mobile device, the enterprise can improve the automation level of the production process and reduce the labor cost, thereby improving the production cost competitiveness.

[0212] In a real industrial scene, the mobile device needs to dock with a docking station (also referred to as a docking object) to carry and transport materials, and therefore has strict requirements on docking accuracy. In the case where the accuracy does not meet the requirements, direct docking may cause damage to the materials or the vehicle body.

[0213] In the process of automatic mobile device docking station, even if the sensors of the automatic mobile device are accurately calibrated, the actual docking accuracy often does not meet the requirements, resulting in material misplacement or failure to place goods. This situation is mainly caused by the following reasons: (1) Multiple automatic mobile devices are not consistent in positioning. In the scenario of docking the same station, although a single automatic mobile device can basically reach the same physical point when repeatedly docking, the actual physical points of different automatic mobile devices are not the same, resulting in deviation of the material placement position of different automatic mobile devices docking the same station, thereby affecting the accuracy of subsequent automatic mobile devices performing tasks on the station; (2) There is an error in the calibration of the docking station. When the automatic mobile device docks the station by positioning, if the map point position of the station deviates from the actual physical position, it will affect the docking accuracy of the automatic mobile device; when the automatic mobile device recognizes the docking station, if there is a deviation between the laid position of the recognition marker and the actual position (such as the left or right deviation of the reflective strip), it will affect the docking accuracy of the automatic mobile device. For example, a current automatic mobile device places a material on a station, and a subsequent automatic mobile device needs to transport the material to another station. If multiple automatic mobile devices are not consistent in positioning, that is, the target docking position (i.e. target point) reached by the first automatic mobile device and the subsequent automatic mobile device docking the station is different, then the material placement position of the first automatic mobile device and the material pickup position of the subsequent automatic mobile device will be inconsistent, which is likely to cause damage to the material or the vehicle body.

[0214] Generally, in order to improve the problem of inconsistent positioning of multiple automatic devices, the conventional technology usually selects multiple points in the global coordinate system and collects positioning data of multiple AGVs at these points. Then, in the multiple AGVs, a certain AGV (referred to as A vehicle) is selected, and its positioning data is taken as reference data. Finally, in the remaining AGVs, a certain AGV (referred to as B vehicle) is selected, and based on the positioning data of B vehicle and the specified constraint condition, B vehicle is calibrated to be consistent with the positioning result of A vehicle. Finally, all the remaining vehicles are traversed, so that the positioning consistency of multiple vehicles is achieved. However, if the positioning data of B vehicle has a deviation, after completing the consistency calibration, all AGVs are consistent with B vehicle, at this time all AGVs have the deviation, which is not conducive to the accurate positioning of AGV. At the same time, the conventional technology does not consider the error in the calibration of the docking station.

[0215] Based on this, another embodiment of the present specification also proposes a positioning compensation method of an automatic mobile device, an electronic device and a computer readable storage medium, which improves the positioning accuracy of the automatic mobile device by simultaneously compensating for the positioning error of the automatic mobile device and the calibration error of the station. For more details, please refer to FIG. 10.

[0216] The execution subject of the positioning compensation method of the automatic mobile device can be a terminal device or a server or other processing device, wherein the terminal device can be a user equipment (UE), a computer, a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some embodiments, the execution subject of the positioning compensation method of the automatic mobile device can also be a positioning compensation system of the automatic mobile device. In some possible implementation manners, the positioning compensation method of the automatic mobile device can be implemented by a processor invoking computer readable instructions stored in a memory. For more details of the positioning compensation system, refer to FIG. 9.

[0217] FIG. 9 is an exemplary module diagram of a positioning compensation system of an automatic mobile device according to some embodiments of the present specification.

[0218] In some embodiments, as shown in FIG. 9, the positioning compensation system 900 of the automatic mobile device can include an error determination module 910, a compensation module 920, a control module 930, and a storage module 940.

[0219] In some embodiments, the error determination module 910 can be configured to obtain a reference first positioning error of the target automatic mobile device docking the target docking position.

[0220] In some embodiments, the error determination module 910 can be further configured to determine a first initial positioning error of the target automatic mobile device at each docking position of the plurality of docking positions, and determine the reference first positioning error based on the first initial positioning error at the plurality of docking positions.

[0221] In some embodiments, the error determination module 910 can be further configured to obtain first perception data collected by the target automatic mobile device at a first actual arrival position related to each docking position after the target automatic mobile device moves to the first actual arrival position, obtain a first error between the first actual arrival position and the docking position, determine a second error between a first perception position based on the first perception data and the docking position, and determine the first initial positioning error of the target automatic mobile device at each docking position according to the first error and the second error.

[0222] In some embodiments, the error determination module 910 can be configured to obtain a reference second positioning error of the target automatic mobile device docking the target docking position.

[0223] In some embodiments, the error determining module 910 can be further configured to: obtain an adjusted target docking position based on the reference first positioning error adjusting the target docking position; obtain second perception data collected by the target automatic moving device at a second actual arrival position related to the target docking position after the target automatic moving device moves to the second actual arrival position; obtain a third error between the second actual arrival position and the target docking position; determine a fourth error between a second perception position based on the second perception data and the target docking position; and determine a second positioning error of the target automatic moving device at the target docking position according to the third error and the fourth error.

[0224] In some embodiments, the compensation module 920 can be configured to compensate the target docking position based on the reference first positioning error and the reference second positioning error to obtain a compensated target docking position.

[0225] In some embodiments, the compensation module 920 can be further configured to: obtain a sum of the reference first positioning error and the reference second positioning error; and determine a sum between the sum of the reference first positioning error and the reference second positioning error and the target docking position as the compensated target docking position.

[0226] In some embodiments, the control module 930 can be configured to control the automatic moving device to move to the compensated target docking position based on the compensated target docking position.

[0227] In some embodiments, the storage module 940 can be configured to store the reference first positioning error and / or the reference second positioning error.

[0228] In some embodiments, the storage module 940 can be configured to store the compensated target docking position.

[0229] For more information about the above modules, please refer to the related descriptions of FIG. 10-FIG. 12.

[0230] It should be noted that the above description of the positioning compensation system 900 of the automatic moving device and its modules is for the convenience of description, and cannot limit the scope of the embodiments. It can be understood that, for those skilled in the art, after understanding the principles of the device, any combination of the modules or connection of the modules to other modules can be made without departing from the principles. In some embodiments, the error determining module 910, the compensation module 920, the control module 930 and the storage module 940 in FIG. 9 can be different modules in one device, or one module can implement the functions of two or more modules described above. For example, the modules can share one storage module, and the modules can also have their own storage modules. Variations such as this are within the scope of protection of the present specification.

[0231] FIG. 10 is an exemplary flowchart of a positioning compensation system, according to some embodiments of the present specification. In some embodiments, as shown in FIG. 10, the flow 1000 includes steps 1010-1040. The flow 1000 can be performed by a terminal device or a server or other processing device.

[0232] At step 1010, a reference first positioning error of a target automatic mobile device to a target docking position is obtained.

[0233] The target automatic mobile device refers to an automatic mobile device for which it is needed to determine whether there is a positioning abnormality. For example, the target automatic mobile device can be an AMR (Autonomous Mobile Robot), an AGV (Automated Guided Vehicle), or the like.

[0234] The target docking position refers to a position (also referred to as a target point or a target point position) at which the target automatic mobile device docks with a docking object. In some embodiments, the target docking position can be the position at which the docking object is located. In some embodiments, the target docking position can be close to the position at which the docking object is located, for example, the distance between the target docking position and the docking object is less than a threshold value, for example, 0.1 meters, 0.05 meters, or the like.

[0235] The reference first positioning error is caused by a positioning device of the target automatic mobile device. For example, the reference first positioning error is a positional deviation between an actual position and a perceived position caused by the positioning device of the target automatic mobile device.

[0236] For example, in an actual docking process, due to the deviation of the target automatic mobile device when positioning, there is a deviation between the actual reaching position of the target automatic mobile device and the perceived position at the actual reaching position, thereby causing the automatic mobile device to fail to reach the target docking position, i.e., there is a deviation between the actual reaching position and the target docking position. For example, the error in the calibration of the positioning device can cause the target automatic mobile device to deviate when positioning.

[0237] In some embodiments, the reference first positioning error of the automatic mobile device (e.g., the target automatic mobile device) relative to the docking position (e.g., the target docking position) refers to the error between the first actual reaching position of the automatic mobile device after positioning to the docking position and the first perceived position of the automatic mobile device.

[0238] In some embodiments, the reference first positioning error can include a first sensory positioning error or a first landmark positioning error. The first sensory positioning error refers to an error between a first perceived position of the automated mobile device at the first actual reached position by autonomous positioning technology and the first actual reached position. The first landmark positioning error refers to an error between a first perceived position of the automated mobile device at the first actual reached position by guided positioning technology (e.g., identifying a landmark (such as a docking object, a two-dimensional code, a magnetic pin, etc.)) and the first actual reached position.

[0239] For example, based on autonomous positioning technology, the target automated mobile device can obtain a map point of the target point (docking object) in a reference map, and autonomously navigate to the map point based on a positioning system (such as SLAM). For another example, based on guided positioning technology, the target automated mobile device can scan and identify a preset landmark (such as a visual marker, a contour, a two-dimensional code, etc.) based on a scanning device (such as an image code reader, a laser code scanner, a depth camera, an RFID reader, etc.), and obtain the position and angle of the landmark relative to the target automated mobile device through image processing and calculation to adjust the pose of the target automated mobile device to reach the target point. In this application, the process of the target automated mobile device reaching the specified target point through autonomous positioning technology is also referred to as positioning to the point; and the process of reaching the specified target point through guided positioning technology is also referred to as identifying to the point.

[0240] For further example, based on autonomous positioning technology, the target automated mobile device (such as an AMR) can obtain a high-precision map (i.e., a reference map) constructed in advance. The reference map includes the position information of each docking object and point. The processor sends the position (e.g., the map point) of the target point to the target automated mobile device: "(x = 15.6, y = 22.1, θ = 0)". The target automated mobile device calculates the position of the target automated mobile device in real time through laser SLAM technology. The processor plans a path from the current position and orientation to the target position and orientation, and controls the target automated mobile device to move. When the target automated mobile device judges that the error between the pose of the target automated mobile device (i.e., the current position and orientation of the target automated mobile device) and the target pose (i.e., the target position and orientation) is within the allowable range (the error is less than a threshold value, which can be preset based on experience), the target automated mobile device considers that it has reached the target point.

[0241] Further for example, based on the guided positioning technology, the processor sends an instruction to the target mobile device: "Please go to find and dock with the A model machine tool" or "Please go to identify and approach the No. 3 two-dimensional code". The target mobile device can move to the vicinity of the docking object (i.e. the target area) based on the pre-planned path corresponding to the docking object in the instruction. After the target mobile device starts the scanning device (such as a laser radar) to scan the docking object to obtain the position information of the docking object (i.e. the target point), it plans a path from the current position to the target point. When the target mobile device reaches the vicinity of the docking object (for example, the edge of the docking object) according to the planned path, it starts the camera to scan and identify the preset marker (such as the visual mark on the machine tool, the outline of the shelf, the two-dimensional code on the ground). Through image processing and calculation, the relative position and angle of the docking object to the target mobile device (i.e. "the target is 0.5 meters in front of the target mobile device and 10 centimeters to the right") are obtained. The target mobile device adjusts the target point according to the relative position and angle (i.e. the pose deviation relative to the target point) for fine adjustment and finally moves to the accurate docking position.

[0242] In some embodiments, the reference first positioning error is pre-set and stored in a storage device (for example, the storage module 940), and the storage device can store the reference first positioning error corresponding to different mobile devices and docking positions. The processor can find and obtain the reference first positioning error corresponding to the target mobile device and the target docking position from the storage device based on the identification of the target mobile device and the target docking position (for example, the pre-set position number).

[0243] In some embodiments, the reference first positioning error of the target mobile device is the same for different docking positions.

[0244] In some embodiments, the reference first positioning error of the target mobile device is different for different docking positions.

[0245] In some embodiments, the reference first positioning error can be obtained based on the following operations.

[0246] For example, if the reference first positioning error is a first sensing positioning error, a plurality of docking objects can be selected, and a map point (i.e., a docking position) of each docking object in the plurality of docking objects can be obtained. Based on the map point of each docking object, the target automatic mobile device can be controlled to move to each docking object to approach or reach a position in the physical world corresponding to the map point (i.e., an ideal docking position, i.e., a docking position or a target point position), to stop at a first actual arrival position. First perception data collected by the target automatic mobile device at the first actual arrival position can be obtained. A first error between the first actual arrival position of the target automatic mobile device and the docking position (i.e., the ideal docking position) can be obtained. A second error can be determined based on the docking position (i.e., the map point) of each target automatic mobile device in the map and the first perception position determined based on the first perception data. A first initial positioning error can be determined based on the first error and the second error. The reference first positioning error of the target automatic mobile device can be determined based on the first initial positioning error corresponding to all docking objects.

[0247] For another example, if the reference first positioning error is a first identification positioning error, the processor determines the reference first positioning error in a manner substantially the same as the above example. The target automatic mobile device can be sent an instruction including docking object related information (e.g., “please find and dock the A model machine tool”). The processor can control the target automatic mobile device to move to the docking object based on the docking object related information instruction. When the target automatic mobile device enters the docking distance range, point cloud data can be obtained by scanning using a laser radar, and the center coordinates of the docking rack (i.e., the docking position or the target point) can be obtained based on the point cloud data. Subsequently, the target automatic mobile device can plan a path from the current position to the docking position and perform path tracking. When the target automatic mobile device reaches the edge or vicinity of the docking position, a camera can be started to scan the identification object (e.g., a positioning code) of the docking object, and the path end point can be modified based on the scanning result. When the target automatic mobile device reaches the first actual arrival position based on the modified path end point, the camera can be used to scan the identification object of the docking object to obtain an image. The processor can obtain the position and angle of the docking object relative to the target automatic mobile device (i.e., “the target is 0.5 meters in front of me and 10 centimeters to the right”) by image processing and calculation, i.e., a second error. The first error can be obtained in the same manner as the first error in the reference first positioning error.

[0248] In some embodiments, the processor can determine a first initial positioning error of the target automatic mobile device at each docking position in a plurality of docking positions, and determine the reference first positioning error based on the first initial positioning error at the plurality of docking positions.

[0249] The first initial positioning error refers to an error between the first actual arrival position and the first perception position of the target automatic mobile device at each docking position.

[0250] In some embodiments, the processor can determine whether the target automated mobile device reaches the docking position (e.g., the reference position information in the reference map) by the autonomous positioning technique.

[0251] In some embodiments, when the target automated mobile device is positioned to the point (i.e., the target automated mobile device self-perceives that it has reached the docking position by the autonomous positioning technique), the target automated mobile device moves to a first actual arrival position associated with each docking position. In some embodiments, the first actual arrival position is close to the docking position, e.g., the distance between the first actual arrival position and the docking position is less than a threshold, 0.1 m, etc. The processor can obtain first perception data collected by the target automated mobile device at the first actual arrival position after the target automated mobile device moves to the first actual arrival position associated with each docking position; obtain a first error between the first actual arrival position and the docking position; determine a second error between a first perception position and the docking position based on the first perception data; and determine a first initial positioning error of the target automated mobile device at each docking position according to the first error and the second error.

[0252] The first actual arrival position refers to the actual position of the target automated mobile device when it docks with the docking object.

[0253] The first perception data includes the position information perceived by the target automated mobile device at the first actual arrival position. The first perception data can be the positioning information of the target automated mobile device at the first actual arrival position obtained by the autonomous positioning technique. For example, the target automated mobile device can collect the first perception data at the first actual arrival position by the positioning sensor (e.g., contact sensor, proximity sensor, etc.) loaded by itself.

[0254] In some embodiments, the processor can obtain the positioning information (i.e., the first perception data) of the target automated mobile device at the first actual arrival position obtained by the autonomous positioning technique by communicating with the target automated mobile device.

[0255] In some embodiments, the first perception data can be the positioning information of the target automated mobile device at the first actual arrival position obtained by the guided positioning technique. For example, the target automated mobile device can determine the first perception data by the detection information obtained by scanning the marker by the scanning device (e.g., scanner, image code reader, laser code scanner, depth camera, RFID reader, etc.).

[0256] The marker can include a barcode, a QR code, an AprilTag, an ArUco, a reflective strip, a reflective plate, etc.

[0257] The detection information refers to information (i.e., the first perception data) related to the pose of the docking object obtained after scanning the marker. For example, the detection information can include the position and orientation of the docking object.

[0258] In some embodiments of the present specification, the first perception data or the second perception data is determined based on the detection information, which eliminates the dependence on a high-precision map or an external reference point. Even if the on-site layout is temporarily adjusted or the ground reflection changes, only the marker needs to be updated, and the entire map does not need to be remeasured. This significantly reduces the deployment and maintenance costs, while improving the adaptation speed of the target automatic mobile device to the dynamic environment and the overall docking accuracy.

[0259] The first error refers to the positional deviation between the first actual arrival position and the docking position. For each of the plurality of docking positions, a first error can be determined accordingly.

[0260] In some embodiments, the first error can be measured by a technician through a ranging tool (such as a laser radar, a range finder, a tape measure, etc.) to measure the distance and orientation between the target automatic mobile device and the docking position. The distance and orientation are the first error between the first actual arrival position and the docking position. For example, the first error can be represented as p err1 =(x err1 ,y err1 ,θ err1 ).

[0261] In other embodiments, the processor can determine the first error based on the point cloud data obtained by the positioning device (such as a laser radar) at the first actual arrival position. For example, the processor can perform point cloud registration on the obtained point cloud data and the preset point cloud data to determine a transformation matrix between the actual obtained point cloud data and the preset point cloud data, and the transformation matrix is the data representation of the first error.

[0262] The second error refers to the positional deviation between the first perception position and the docking position. For each of the plurality of docking positions, a second error can be determined accordingly.

[0263] It should be noted that when the target automatic mobile device docks by positioning to a point, whether the target automatic mobile device is positioned to the point is mainly determined according to the data provided by the sensor of the target automatic mobile device. When the data of the sensor shows that the target automatic mobile device has reached the docking position, it is considered that the target automatic mobile device is positioned to the point. There can be a certain deviation between the data of the sensor and the docking position, i.e., the data of the sensor does not need to be completely consistent with the docking position to confirm that the target automatic mobile device is positioned to the point. The deviation is generally less than 1 centimeter. Therefore, the second error between the first perception position and the docking position needs to be calculated in the present embodiment, and the second error is often referred to as a nominal to-point error.

[0264] The first perceived position is a positioning position obtained by the target automatic moving device through autonomous positioning technology or guided positioning technology.

[0265] In some embodiments, the processor can determine a second error based on a difference between the first perceived position and the docking position. The first perceived position can be represented as p loc1 =(x loc1 ,y loc1 ,θ loc1 ), the docking position can be represented as p map , and the second error can be represented as p nor_err1 =p loc1 -p map .

[0266] In some embodiments, the processor can determine a first initial positioning error based on the first error and the second error.

[0267] For example, the processor can take a difference between the first error and the second error as the first initial positioning error. The first error can be represented as p err1 , the second error can be represented as p nor_err1 , and the first initial positioning error can be represented as p cons_i =p err1 -p nor_err1 .

[0268] For another example, the processor can also perform weighted sum processing on the first error and the second error to obtain the first initial positioning error of the target automatic moving device at the docking position, which can be represented as p cons_i =a*p err1 +b*p nor_err1 .

[0269] For each of the N docking positions, the above steps are repeated to obtain the first initial positioning error of the target automatic moving device relative to each docking position, which can be represented as [p cons_1 ,p cons_2 ,...,p cons_N ].

[0270] In some embodiments, the processor can first select a plurality of docking objects, for each of which, a map point of the docking object is obtained; based on the map point of the docking object, control the target automatic mobile device to move to a docking position of the docking object (i.e., the map point of the docking object) or a position in the physical world corresponding to the map point (i.e., an ideal docking position) to stop moving at a first actual arrival position; obtain first perception data collected by the target automatic mobile device at the first actual arrival position; and obtain a first error and a second error of the target automatic mobile device, and take a difference between the first error and the second error as a first initial positioning error of one docking object. Wherein the map point of the docking object is consistent with the position of the docking object in the physical world (i.e., the ideal docking position).

[0271] In some embodiments, the processor can determine a mean value of the plurality of first perception data based on the first perception data of the plurality of target automatic mobile devices at the docking position, and if a difference between the mean value and the map point of the docking position is less than a pre-set threshold, it is considered that the map point of the docking position and its position in the physical world are relatively consistent, otherwise the map point of the docking object position is modified to its position in the physical world.

[0272] In some embodiments, when the target automatic mobile device performs a task, a target point position (i.e., a map point) corresponding to the task is issued, and then the target automatic mobile device can move to a position in the physical world corresponding to the target point position (i.e., an ideal docking position) based on a path between the target point position and the current position. However, the target automatic mobile device generally cannot reach the ideal docking position, but will reach a position close to the ideal docking position (i.e., a first actual arrival position), which can correspond to a first perception position. The processor can issue a task of docking the docking object to each target automatic mobile device, and the task includes the target point position (i.e., the map point or the target docking position). The target automatic mobile device can use sensors and maps to autonomously navigate, plan and optimize paths to the docking object.

[0273] In some embodiments of the present specification, determining the first initial positioning error based on the first error and the second error can decouple and quantify "robot body error" and "sensor-to-point nominal error" at one time, so that in the subsequent calibration stage, a cleaner and more representative error estimate can be obtained by calculating only once, reducing the number of repeated calibrations, and significantly improving the overall calibration efficiency and accuracy.

[0274] In some embodiments, the processor can determine the first initial positioning error of the target automatic mobile device at each of the plurality of docking objects based on the first error between the first actual arrival position and the docking position. For example, the processor can determine the first error as the first initial positioning error, or can determine a preset multiple of the first error as the first initial positioning error. In some embodiments, the processor can determine the first initial positioning error of the target automatic mobile device at each of the plurality of docking objects based on the second error between the first perceived position and the docking position. Similarly, the processor can determine the second error as the first initial positioning error, or can determine a preset multiple of the second error as the first initial positioning error.

[0275] In some embodiments, the processor can determine the reference first positioning error based on the first initial positioning errors of the target automatic mobile device at the plurality of docking objects in various ways. For example, the processor can determine a central tendency value (such as a mean, a maximum value, a median, or a minimum value, etc.) of the first initial positioning errors of the target automatic mobile device at the plurality of docking objects as the reference first positioning error. For example, the processor can count the error sum of the first initial positioning errors corresponding to each docking position and the total number of positions of each docking position; and determine the ratio between the error sum and the total number of positions as the reference first positioning error. For another example, the processor can also perform weighted summation processing on the first initial positioning errors corresponding to each docking position to obtain a weighted error sum; and determine the ratio between the weighted error sum and the total number of positions as the reference first positioning error.

[0276] After obtaining the reference first positioning error, the processor saves the reference first positioning error in the memory of the device itself for subsequent use.

[0277] In some embodiments of the present specification, determining the reference first positioning error based on the first initial positioning error not only effectively smooths the single-point random error and the local environmental noise, but also significantly improves the robustness and global representativeness of error estimation; at the same time, it provides a unified compensation benchmark for all subsequent tasks, reduces the workload of point-by-point calibration, realizes "one-time calibration, full-scene reuse", and thus greatly reduces the deployment and maintenance costs under the premise of ensuring accuracy.

[0278] In some embodiments, the processor can determine whether the target automatic mobile device arrives at the docking position by recognizing the point. Recognizing the point refers to achieving the docking of the target automatic mobile device by recognizing the marker on the docking position. Recognizing the point requires that the marker is laid on the docking position, and the position information of the marker is determined as the position information of the docking position, wherein the position information of the docking position can be obtained by the target automatic mobile device recognizing the marker, which can be represented as p tar = (x tar ,y tar ,θ tar ).

[0279] In some embodiments, the first perception data can be determined based on detection information obtained by the target automatic mobile device scanning the marker. The processor can obtain detection information for the marker; perform identification processing according to the detection information to obtain a docking position of the target automatic mobile device. The target automatic mobile device is controlled to dock at the docking position by identifying the point. The position information of the docking position can be the positioning information of the marker; after docking is completed, the first perception data of the target automatic mobile device is obtained; and the first initial positioning error of the target automatic mobile device is determined according to the first identification error and the second identification error. It should be noted that the identification of the point does not need to use the map positioning information of the target docking position, and therefore whether the map positioning information of the target docking position is accurate does not affect the process. The determination of the first error is the same as the determination of the first error of the positioning to the point by the autonomous navigation technology.

[0280] In some embodiments, the processor can complete multi-vehicle consistency calibration after determining the reference first positioning error of the plurality of target automatic mobile devices relative to the target docking position, to ensure that each target automatic mobile device can accurately reach the target docking position.

[0281] Step 1020, obtaining a reference second positioning error of the target automatic mobile device relative to the target docking position.

[0282] The reference second positioning error is caused by the calibration of the target docking position. For example, the reference second positioning error is the pose error between the actual position and the perception position caused by the calibration of the target docking position.

[0283] In the actual docking process, due to the deviation of the calibration of the target docking position, the actual arrival position of the target automatic mobile device deviates from the perception position of the target docking position at the actual arrival position, so that the automatic mobile device cannot reach the target docking position, that is, there is a deviation between the actual arrival position and the target docking position.

[0284] In some embodiments, the reference second positioning error of the automatic mobile device (for example, the target automatic mobile device) relative to the docking position (for example, the target docking position) refers to the error between the second actual arrival position of the automatic mobile device after positioning to the docking position and the second perception position of the automatic mobile device.

[0285] In some embodiments, the reference second positioning error can include a second sensor positioning error or a second identification positioning error. The second sensor positioning error refers to an error between a second perceived position at the second actual arrival position sensed by the autonomous mobile device at the second actual arrival position through autonomous positioning technology and the second actual arrival position. The second identification positioning error refers to an error between a second perceived position at the second actual arrival position determined by the autonomous mobile device through guided positioning technology (e.g., identifying markers (such as docking objects, two-dimensional codes, magnetic nails, etc.)) and the second actual arrival position.

[0286] In some embodiments, the reference second positioning error is preset and stored in a storage device (e.g., the storage module 940), which can store reference second positioning errors corresponding to different autonomous mobile devices and different docking positions. The processor can find and acquire the reference second positioning error corresponding to the target autonomous mobile device and the target docking position from the storage device based on the identification (e.g., a preset position number) of the target mobile device and the target docking position. For example, the reference second positioning error can be stored in a reference map. The reference second positioning error corresponds to the position information of different docking objects in the reference map, and each docking object can correspond to a reference second positioning error.

[0287] In some embodiments, the reference second positioning errors of different autonomous mobile devices for the same docking position are the same.

[0288] In some embodiments, the reference second positioning errors of different autonomous mobile devices for the same docking position are different.

[0289] In some embodiments, the reference second positioning error can be acquired based on the following operations.

[0290] For example, the processor can directly acquire the reference second positioning error through the storage device.

[0291] In some embodiments, the processor can first compensate the target docking position based on the reference first positioning error to obtain a candidate compensation position, and then acquire the reference second positioning error of the target autonomous mobile device relative to the candidate compensation position. The candidate compensation position refers to a position obtained by compensating the target docking position based on the reference first positioning error. In the scenario of identifying a point, the processor can read the reference first positioning error of the target autonomous mobile device identifying the point from the storage device and compensate the target docking position using the reference first positioning error to obtain the candidate compensation position, which can be represented as

[0292] In some embodiments, the processor can obtain an adjusted target docking position (may also be referred to as a candidate compensation position) based on the reference first positioning error adjustment of the target docking position; obtain second perception data collected by the target automatic mobile device at a second actual arrival position related to the adjusted target docking position (i.e., the candidate compensation position) after the target automatic mobile device moves to the second actual arrival position; obtain a third error between the second actual arrival position and the target docking position; determine a fourth error between a second perception position determined based on the second perception data and the target docking position; and determine a reference second positioning error of the target automatic mobile device at the target docking position according to the third error and the fourth error.

[0293] In some embodiments, the processor can adjust the target docking position in multiple ways based on the reference first positioning error and the target docking position. For example, the processor can translate or rotate the position coordinates of the target docking position by an equal amount in the opposite direction of the reference first positioning error to obtain an adjusted target docking position.

[0294] The second actual arrival position refers to the actual position of the target automatic mobile device when it arrives at the target docking position compensated based on the reference first positioning error (i.e., the candidate compensation position).

[0295] The second perception data includes position information perceived by the target automatic mobile device at the second actual arrival position. The second perception data can be positioning information of the target automatic mobile device at the second actual arrival position obtained by autonomous positioning technology. For example, the target automatic mobile device can collect the second perception data at the second actual arrival position through a positioning sensor (e.g., a contact sensor, a proximity sensor, etc.) loaded on itself.

[0296] In some embodiments, the processor can obtain the positioning information (i.e., the second perception data) of the target automatic mobile device at the second actual arrival position obtained by autonomous positioning technology by communicating with the target automatic mobile device.

[0297] In some embodiments, the second perception data can be positioning information of the target automatic mobile device at the second actual arrival position obtained by guided positioning technology. For example, the target automatic mobile device can determine the second perception data through detection information obtained by scanning a marker through a scanning device (e.g., an image code reader, a laser code scanner, a depth camera, an RFID reader, etc.). For more details, please refer to the relevant description above.

[0298] The third error refers to the positional deviation between the second actual arrival position and the target docking position.

[0299] In some embodiments, the processor can determine a difference between the second actual arrival position and the target docking position as a third error. In other embodiments, the distance and the orientation between the target automatic moving device and the docking position can also be measured in the field. The distance and the orientation are the third error between the second actual arrival position and the docking position. For example, the third error can be represented as p err2 =(x err2 ,y err2 ,θ err2 ).

[0300] The fourth error refers to the positional deviation between the second perceived position determined by the second perception data and the target docking position. In some embodiments, the processor can determine a difference between the second perceived position determined by the second perception data and the target docking position as the fourth error. The second perceived position can be represented as p loc2 =(x loc2 ,y loc2 ,θ loc2 ), and the fourth error can be represented as

[0301] In some embodiments, the second error can be determined based on the first perception data and reference position information of the docking position in the reference map; and the fourth error can be determined based on the second perception data and reference position information of the target docking position in the reference map.

[0302] The reference map refers to a map of a transportation scene in which the target automatic moving device is located, which is preset.

[0303] In some embodiments, the reference map can be preset by the processor by default.

[0304] The reference position information refers to the position information of the docking position in the reference map.

[0305] In some embodiments, the processor can determine the relative relationship between the reference position information and the boundary of the reference map based on the relative relationship between the docking position of the docking object in the physical world and the boundary of the scene, and determine the reference position information based on the relative relationship.

[0306] In some embodiments, the processor can determine a difference between the first perceived position determined by the first perception data and the reference position information as the second error, and determine a difference between the second perceived position determined by the second perception data and the reference position information as the fourth error.

[0307] In some embodiments of the present specification, the second error and the fourth error are determined based on the reference position information, which can be used to calibrate multiple target automatic mobile devices and all docking objects at the same time with the same static true value map, thereby avoiding the cumulative error caused by inconsistent maps; simplifying data management and ensuring high consistency between multiple vehicles and multiple stations, thereby significantly improving overall accuracy and operation efficiency in large-scale deployment.

[0308] In some embodiments, the processor can determine the reference second positioning error in multiple ways based on the third error and the fourth error. For example, the processor determines the reference second positioning error according to the third error and the fourth error after determining the third error and the fourth error. In some embodiments, the difference between the third error and the fourth error can be determined as the reference second positioning error, which can be represented as r rec = p err2 -p nor_err2 In other embodiments, the first error and the second error can also be weighted and summed to obtain the reference second positioning error, which can be represented as p cons_i = c * p err2 +d * p nor_err2 .

[0309] After obtaining the reference second positioning error, the second positioning error is saved in the map for subsequent use. When the docking position is calibrated by positioning to the point, the obtained reference second positioning error can be represented as r pos .

[0310] In other embodiments, the processor can also determine the second positioning error according to the third error. For example, the third error is determined as the second positioning error, or the third error can be processed to obtain the second positioning error. In other embodiments, the processor can also determine the second positioning error according to the fourth error. For example, the fourth error is determined as the second positioning error, or the fourth error can be processed to obtain the second positioning error.

[0311] In some embodiments, the processor can determine multiple second initial positioning errors of multiple target automatic mobile devices at the target docking position; and determine the reference second positioning error based on the multiple second initial positioning errors.

[0312] The processor can determine the reference second positioning error in multiple ways based on the multiple second initial positioning errors. For example, the processor can determine a central tendency value (such as a mean value, a maximum value, a median value, or a minimum value, etc.) of the multiple second initial positioning errors as the reference second positioning error. For example, the processor can count a total error sum of the multiple second initial positioning errors and a total number of the automatic mobile devices; and determine a ratio between the total error sum and the total number of the automatic mobile devices as the reference second positioning error. For another example, the processor can also perform a weighted sum processing on the second initial positioning errors to obtain a weighted error sum; and determine a ratio between the weighted error sum and the total number of the automatic mobile devices as the reference second positioning error.

[0313] The second initial positioning error refers to an error between a second actual arrival position and a second perceived position of each automatic mobile device at the target docking position.

[0314] The processor can first compensate the target docking position based on the reference first positioning error of each automatic mobile device to obtain a candidate compensation position, and then obtain the reference second positioning error of each automatic mobile device relative to the candidate compensation position. For example, after each automatic mobile device moves to a second actual arrival position related to the adjusted target docking position (i.e., the candidate compensation position), the second perception data collected by each automatic mobile device at the second actual arrival position is obtained; a third error between the second actual arrival position and the target docking position is obtained; a fourth error between the second perceived position and the target docking position is determined based on the second perception data; and the second initial positioning error of each automatic mobile device at the target docking position is determined according to the third error and the fourth error. The determination of the third error and the fourth error can refer to the determination of the third error and the fourth error described anywhere in the present application.

[0315] In some embodiments of the present specification, determining the reference second positioning error based on the third error and the fourth error can accurately decouple the "station / map residual error" and the "sensor nominal error" at one time, without additional hardware, and only two rounds of docking are needed to simultaneously correct the target automatic mobile device body error and the station map error; making the error estimation more robust and globally consistent, and subsequent tasks can directly call to achieve millimeter-level repeat accuracy; significantly reducing the number of on-site manual re-measurements and reducing downtime maintenance costs.

[0316] In step 1030, the target docking position is compensated based on the reference first positioning error and the reference second positioning error to obtain a compensated target docking position.

[0317] The compensated target docking position (which can also be referred to as a target compensated position) refers to a position obtained by compensating the target docking position based on the reference first positioning error and the reference second positioning error. The purpose of compensation is to enable the target automatic mobile device to reach the target docking position more accurately.

[0318] It should be noted that, in the docking process of the automatic mobile device, the error includes two parts, the first part is the docking error of the automatic mobile device itself, and the second part is the positioning error of the docking station. Therefore, the embodiment solves the two parts of errors by a two-step approach. The reference first positioning error is used to solve the docking error of the automatic mobile device itself, and the reference second positioning error is used to solve the positioning error of the docking position. In the subsequent task processing process of the target automatic mobile device, the target automatic mobile device can obtain the reference first positioning error of the target automatic mobile device and the reference second positioning error of the target docking position to perform positioning compensation, and obtain the final target compensated position.

[0319] In some embodiments, the processor can first compensate the target docking position based on the reference first positioning error to obtain a candidate compensated position, and then compensate the candidate compensated position based on the reference second positioning error to obtain the target compensated position. For example, if the reference first positioning error is a right deviation of 5° and the reference second positioning error is an upward deviation of 10°, then when the target automatic mobile device docks, the angle of the target docking position is compensated 5° to the left to obtain a candidate compensated position, and then 10° downward to obtain a target compensated position.

[0320] In some embodiments, the processor can obtain the sum of the reference first positioning error and the reference second positioning error; and determine the sum of the reference first positioning error and the reference second positioning error and the target docking position as the compensated target docking position.

[0321] In some embodiments, the processor can obtain the sum of the reference first positioning error and the reference second positioning error after determining the reference second positioning error of each docking position; and determine the sum of the reference first positioning error and the reference second positioning error as the target compensated position. For example, when the target automatic mobile device receives a docking task, if it is a positioning point, the map positioning information p map The compensation is If it is an identified point, the positioning information p tar The compensation is

[0322] In some embodiments of the present specification, the reference first positioning error of the target automatic mobile device and the reference second positioning error of the target docking position are used to jointly compensate the target docking position, so that high-precision docking of the target automatic mobile device and the target docking position can be achieved.

[0323] The target compensation position is the position of the target docking position compensated by the reference first positioning error and the reference second positioning error. It should be noted that, during the docking of the target automatic mobile device, the error includes two parts, the first part is the docking error of the target automatic mobile device itself, and the second part is the positioning error of the docking station. Therefore, the present embodiment solves the two parts of error by a two-step approach. The reference first positioning error is used to solve the docking error of the automatic mobile device itself, and the reference second positioning error is used to solve the positioning error of the docking station. In the subsequent task processing process of the automatic mobile device, the automatic mobile device can obtain the first positioning error of the automatic mobile device and the second positioning error of the docking station for positioning compensation to obtain the final target compensation position.

[0324] It can be seen that, in the automatic mobile device positioning compensation method of the present application, the first positioning error is first determined by the first actual positioning information of the automatic mobile device docking the target docking position and the first target positioning information; the target docking position is compensated according to the first positioning error to obtain the compensated target docking position; then the automatic mobile device docks the compensated target docking position, and the second positioning error is determined by the second actual positioning information after the automatic mobile device docking is completed and the second target positioning information; then the target docking position is compensated according to the reference first positioning error and the reference second positioning error to obtain the target compensation position. Thus, the positioning error of the automatic mobile device is determined by two dockings, which not only solves the positioning error of the automatic mobile device itself, but also solves the calibration error of the docking station, thereby ensuring the docking accuracy of the target automatic mobile device.

[0325] Step 1040, based on the compensated target docking position, controlling the automatic mobile device to move to the target docking position.

[0326] After the compensated target docking position is determined, a task of docking the target docking position is issued to the target automatic mobile device to control the target automatic mobile device to move to the compensated target docking position.

[0327] The positioning compensation method of the automatic mobile device in the embodiment of the present specification first determines the reference first positioning error of the target automatic mobile device relative to the target docking position and the reference second positioning error of the target automatic mobile device relative to the target docking position; and then compensates the target docking position according to the reference first positioning error and the reference second positioning error to obtain the target compensation position. Thus, the positioning error of the target automatic mobile device is determined through two times of docking, which can not only solve the positioning error of the target automatic mobile device itself, but also solve the calibration error of the docking position, thereby ensuring the docking accuracy of the automatic mobile device.

[0328] To ensure the positioning information accuracy of the target docking position, the present embodiment provides a way of verifying the positioning information of the target docking position. In some embodiments, the processor can determine whether the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world satisfy a preset requirement; and in response to determining that the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world do not satisfy the preset requirement, adjusting the reference position information of the target docking position in the reference map until the preset requirement is satisfied.

[0329] The reference position information refers to the relative position information of the target docking position on the reference map. The actual position information refers to the absolute position information of the target docking position in the physical world.

[0330] The reference position information of the target docking position in the reference map can be determined in advance. For example, when the docking position is set, the measured positioning information of the docking position is measured to determine the reference position information of the docking position in the reference map.

[0331] In some embodiments, the processor can select a docking position with good positioning environment as the target docking position in the present embodiment, and obtain the reference position information of the target docking position in the reference map.

[0332] In some embodiments, the processor can measure the measured positioning information of the target docking position to obtain the actual position information of the target docking position in the physical world.

[0333] In some embodiments, the preset requirement can be that the difference between the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world is less than a preset threshold. When the difference between the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world is less than the preset threshold, it can be considered that the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world satisfy the preset requirement. Otherwise, the preset requirement is not satisfied.

[0334] In some embodiments, the preset requirement can also be that a difference between the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world is within a preset range. When the difference between the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world is within the preset range, it can be considered that the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world meet the preset requirement. Otherwise, the preset requirement is not met.

[0335] The preset threshold, the preset range can be a system default value, an empirical value, a human pre-set value, etc. or any combination thereof, which can be set according to actual needs, and the present specification does not limit this.

[0336] The above description of the preset requirement is only for illustrative purposes and is not intended to limit the scope of the present specification.

[0337] In some embodiments, the processor can acquire third perception data collected by the plurality of automatic mobile devices at the target docking position; determine a third perception position perceived by the plurality of automatic mobile devices at the target docking position based on the third perception data; and determine whether the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world meet the preset requirement based on whether a deviation between the third perception position and the reference position information of the target docking position in the reference map is within a preset range.

[0338] The third perception data contains perception positioning information of the plurality of automatic mobile devices at the target docking position. The acquisition method of the third perception data is similar to that of the first perception data and the second perception data, which will not be described here.

[0339] The third perception position can be a position determined based on the third perception data of the plurality of automatic mobile devices at the target docking position. For the same target docking position, the processor can calculate the average value of the perception positioning information of each automatic mobile device at the target docking position, and take the average value as the third perception position.

[0340] In some embodiments, the number of automatic mobile devices can be greater than or equal to 20, that is, 20 automatic mobile devices are used to collect perception positioning information at each docking position.

[0341] FIG. 11 is another exemplary flowchart of a positioning compensation method according to some embodiments of the present specification. In some embodiments, as shown in FIG. 11, the flow 1100 includes steps 1110-1140. The flow 1100 can be executed by a terminal device or a server or other processing device.

[0342] In step 1110, in response to detecting the at least one automatic mobile device docking completion, a first positioning error is determined according to the acquired first actual positioning information of the automatic mobile device and the first target positioning information, the first target positioning information representing the positioning information when the automatic mobile device is at the at least one target docking position.

[0343] The automatic mobile device is configured to receive a task instruction to perform a corresponding task. For example, the automatic mobile device can be an AMR (Autonomous Mobile Robot), an AGV (Automated Guided Vehicle), or the like.

[0344] The first actual positioning information is the positioning information of the automatic mobile device after docking completion. For example, the first actual positioning information can include, but is not limited to, perception positioning information, measured positioning information, or the like in the automatic mobile device. The perception positioning information can be the positioning information perceived by the sensor carried by the automatic mobile device. The measured positioning information can be the positioning information obtained by measuring the positioning of the automatic mobile device in the field.

[0345] The first target positioning information is the expected docking position of the automatic mobile device, i.e., the actual position information when the automatic mobile device is at the at least one target docking position. For example, a reference mobile device can be determined, controlled to move to the target docking position, and then the positioning information of the reference mobile device is acquired as the first target positioning information. In the actual docking process, due to various reasons of the automatic mobile device, the first actual positioning information of the automatic mobile device deviates from the first target positioning information, resulting in that the automatic mobile device cannot accurately place the material. Therefore, the embodiment compensates the positioning of the automatic mobile device according to the deviation between the first actual positioning information and the first target positioning information, so as to improve the docking accuracy of the automatic mobile device.

[0346] In some application scenarios, the positioning information of the reference mobile device at the target docking position can be regarded as the map positioning information of the target docking position, so that the map positioning information of the target docking position can be determined as the positioning information of the target docking position. Of course, the reference mobile device can also be measured in the field to determine the positioning information of the target docking position. In another application scenario, a marker can be arranged on the docking station to measure the position of the marker to determine the positioning information of the target docking position.

[0347] The first positioning error is used to measure the deviation of the automatic mobile device after docking the docking target docking position. Exemplarily, the first positioning error of the automatic mobile device can be determined by the first actual positioning information of the automatic mobile device and the first target positioning information. In some embodiments, if there is a deviation between the first actual positioning information and the first target positioning information, indicating that the positioning of the automatic mobile device is inaccurate, it is necessary to jump to step 1120 for positioning compensation; if there is no deviation between the first actual positioning information and the first target positioning information, indicating that the positioning of the automatic mobile device is accurate, no positioning compensation is needed. In other embodiments, if the first positioning error between the first actual positioning information and the first target positioning information is greater than a preset error, jump to step 1120 for positioning compensation; if the first positioning error between the first actual positioning information and the first target positioning information is less than or equal to the preset error, no positioning compensation is needed.

[0348] In step 1120, the target docking position is compensated according to the first positioning error to obtain a compensated target docking position.

[0349] The compensated target docking position refers to the position obtained by compensating the target docking position according to the first positioning error. The purpose of compensation is to enable the automatic mobile device to reach the target docking position more accurately. For example, if the first positioning error of the automatic mobile device is 5° deviation to the right, then when the automatic mobile device docks, the angle of the target docking position is compensated 5° to the left to obtain the compensated target docking position.

[0350] In step 1130, during the docking of the automatic mobile device at the compensated target docking position, in response to detecting that the automatic mobile device has docked, the second positioning error is determined according to the obtained second actual positioning information and the second target positioning information of the automatic mobile device, and the second target positioning information represents the positioning information of the automatic mobile device at the compensated target docking position.

[0351] The second actual positioning information is the positioning information of the automatic mobile device after docking the compensated target docking position. Exemplarily, the second actual positioning information can include but is not limited to perception positioning information, measured positioning information, etc. in the automatic mobile device. The perception positioning information can be the positioning information perceived by the sensor carried by the automatic mobile device. The measured positioning information can be the positioning information obtained by measuring the positioning of the automatic mobile device in the field.

[0352] The second target positioning information is target docking position of the automatic mobile device after docking compensation, that is, the positioning information when the automatic mobile device is in the target docking position after compensation. In the actual docking process, due to the deviation of the docking station and other reasons, the second actual positioning information of the automatic mobile device deviates from the second target positioning information, so that the automatic mobile device cannot accurately place the material. Therefore, the embodiment compensates the positioning of the automatic mobile device through the deviation between the second actual positioning information and the second target positioning information, so as to improve the docking accuracy of the automatic mobile device.

[0353] The second positioning error is used to measure the deviation of the target docking position of the automatic mobile device after docking compensation. Exemplarily, the second positioning error of the automatic mobile device can be determined through the second actual positioning information and the second target positioning information of the automatic mobile device. In some embodiments, if there is a deviation between the second actual positioning information and the second target positioning information, it means that the automatic mobile device cannot accurately dock, and it is necessary to jump to step 1120 for positioning compensation; if there is no deviation between the second actual positioning information and the second target positioning information, it means that the automatic mobile device can accurately dock, and it is not necessary to perform positioning compensation. In other embodiments, if the second positioning error between the second actual positioning information and the second target positioning information is greater than a preset error, the positioning compensation is performed in step 1120; if the second positioning error between the second actual positioning information and the second target positioning information is less than or equal to the preset error, the positioning compensation is not necessary.

[0354] In step 1140, the target docking position of the automatic mobile device is compensated according to the first positioning error and the second positioning error, to obtain a target compensation position.

[0355] The target compensation position is the position of the target docking position after being compensated by the first positioning error and the second positioning error. It should be noted that in the docking process of the automatic mobile device, the error includes two parts, the first part is the docking error of the automatic mobile device itself, and the second part is the positioning error of the docking station. Therefore, the embodiment solves the two parts of errors in two steps. The first positioning error is used to solve the docking error of the automatic mobile device itself, and the second positioning error is used to solve the positioning error of the docking station. In the subsequent task processing process of the automatic mobile device, the automatic mobile device can obtain the first positioning error of the automatic mobile device and the second positioning error of the docking station for positioning compensation, to obtain the final target compensation position.

[0356] Therefore, the positioning error of the automatic mobile device is determined through two dockings, which not only solves the positioning error of the automatic mobile device itself, but also solves the calibration error of the docking station, so as to ensure the docking accuracy of the automatic mobile device.

[0357] To elaborate the positioning compensation method of the automatic mobile device provided by some embodiments of the present specification, a further illustration is made by means of a flowchart shown in FIG. 12.

[0358] FIG. 12 is another exemplary flowchart of the positioning compensation method of the automatic mobile device according to some embodiments of the present specification. In some embodiments, the flow 1200 can be executed by a terminal device or a server or other processing device. As shown in FIG. 12, the flow 1200 includes the following contents:

[0359] First of all, it needs to be pointed out that the present embodiment includes two stages. The first stage is the multi-vehicle consistency calibration. Since in the scenario of docking the same docking station, although a single automatic mobile device can reach the same physical point when repeatedly docking, the actual physical points reached by different automatic mobile devices are not the same, which affects the docking accuracy. The second stage is the docking station calibration. When the automatic mobile device docks the docking station by positioning to the point, if there is a large deviation between the map point of the docking station and the actual physical position, it will also affect the docking accuracy of the automatic mobile device. When the automatic mobile device docks the docking station by recognizing the point, if there is a deviation between the laying position of the marker and the actual docking position, such as the reflective strip being left or right, etc., it will also affect the docking accuracy of the automatic mobile device.

[0360] Based on this, in the first stage, the first positioning error of the automatic mobile device is obtained. It includes: first determining the first target positioning information by the reference mobile device; for each automatic mobile device, controlling the automatic mobile device to dock from any one of the docking stations for multi-vehicle consistency calibration, and obtaining the first measured positioning information and the first perceived positioning information of the automatic mobile device after the docking is completed; calculating the first error between the first measured positioning information and the first target positioning information, and calculating the second error between the first perceived positioning information and the first target positioning information; determining the difference between the first error and the second error as the first initial positioning error of the automatic mobile device; and obtaining the first positioning error of the automatic mobile device by summing up the first initial positioning error of the automatic mobile device docking each station.

[0361] In the second stage, the second positioning error of the docking station is determined. It includes selecting any automatic mobile device calibrated by consistency to dock the docking station that needs to be calibrated, compensating the target docking position corresponding to the docking station by the first positioning error before docking, and controlling the automatic mobile device to dock the compensated target docking position; obtaining the second measured positioning information and the second perceived positioning information of the automatic mobile device after the docking is completed; and determining the third error by the second measured positioning information and the second target positioning information, and determining the fourth error by the second perceived positioning information and the second target positioning information; and determining the second positioning error of the docking station according to the third error and the fourth error.

[0362] The first positioning error of the automatic mobile device and the second positioning error of the docking station are used to jointly compensate the target docking position, so that high-precision docking of the automatic mobile device is realized.

[0363] One or more embodiments of the present specification provide a positioning anomaly detection device of an automatic mobile device, the device comprising at least one processor and at least one memory; the at least one memory is configured to store computer instructions; the at least one processor is configured to execute at least part of the computer instructions to implement an anomaly detection method.

[0364] One or more embodiments of the present specification provide a computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, a positioning anomaly detection method of an automatic mobile device is implemented. In some embodiments, the computer readable storage medium stores program instructions executable by the processor. The program instructions are used to implement the steps in any of the positioning anomaly detection method embodiments of the automatic mobile device described above.

[0365] One or more embodiments of the present specification provide a positioning compensation device of an automatic mobile device, the device comprising at least one processor and at least one memory; the at least one memory is configured to store computer instructions; the at least one processor is configured to execute at least part of the computer instructions to implement a positioning compensation method of an automatic mobile device.

[0366] One or more embodiments of the present specification provide a computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, a positioning compensation method of an automatic mobile device is implemented.

[0367] In some embodiments, the device provided by the embodiments of the present specification has functions or includes modules that can be used to execute the methods described in the above method embodiments, and the specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.

[0368] The above description of each embodiment tends to emphasize the differences between each embodiment, and the same or similar parts can be mutually referred to. For the sake of brevity, it will not be repeated here.

[0369] In several embodiments provided in the present specification, it should be understood that the disclosed method and device can be implemented in other manners. For example, the division of the apparatus embodiments is merely illustrative, and for example, the division of the modules or units can be other division manners, for example, one unit can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections can be indirect couplings or communication connections through some interfaces, devices or units, and can be in electrical, mechanical or other forms.

[0370] In addition, each function unit in each embodiment of the present specification can be integrated into a processing unit, or each unit can exist physically independently, or two or more units can be integrated into one unit. The integrated unit can be implemented in the form of hardware or in the form of a software function unit.

[0371] If the integrated unit is implemented in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such an understanding, the technical solutions of the present application essentially or the part that makes a contribution to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform all or part of the steps of the methods in the various embodiments of the present application. The foregoing storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0372] The above has described the basic concepts, and it is obvious that the above detailed disclosure is merely an example and does not limit the present specification. Although it is not explicitly stated herein, those skilled in the art can make various modifications, improvements and corrections to the present specification. Such modifications, improvements and corrections are suggested in the present specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of the present specification.

[0373] Also, the use of "a" or "an" or "the" or "at least one" or "one or more" or "one or more instances" throughout the specification try to convey a similar meaning as the term "one or more" unless the context clearly dictates otherwise. The terms "comprising," "including," "containing," and "having" are intended to be open-ended terms. Likewise, the term "comprises" is synonymous with "includes" or "contains" for purposes of the specification and claims. Therefore, use of these terms is not intended to limit the scope of the disclosure to the specific embodiments discussed. Furthermore, the terms "first," "second," "third," and the like are used merely to distinguish one element from another, and do not require that the elements be in any particular order. It is also noted that various implementations of the disclosure have been described as comprising, including, containing, comprising, having or any other comparable term. However, the term "comprising" is used herein to mean the open-ended term "including but not limited to," and thus should be interpreted to cover the terms "consisting of" and "consisting essentially of" and their grammatical equivalents. In some embodiments, the term "consisting essentially of" is used to mean the open-ended term "including, but not limited to, and one or more additional elements of the same, different kind, unless the context clearly indicates otherwise."

[0374] Furthermore, the order of the processing elements and sequences described in this specification are not intended to be construed as a limitation, unless specifically stated otherwise. Additionally, the use of "a" or "an" are employed to describe elements and do not exclude the presence of more than one of such elements. Moreover, although the terms "first," "second," and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are generally only used to distinguish one element from another. Thus, a reference to the first element could be a distinction from a second element or a third element, which in some instances can be the same. The terminology used herein is not intended to be interpreted in any particular manner except as provided in the following sections. For example, the use of the terms "including," "comprising," or "having" and variations thereof herein is meant to encompass the items listed thereafter and equivalents thereof as well as additional items. Unless specified otherwise, like reference numerals designate corresponding parts throughout the description of the specification.

[0375] Similarly, it is to be noticed that the term "comprising", used in the description, is not used in the sense restricting itself to the features of the specific embodiments described. It is used in the sense of "including" and in the sense of "consisting of" and in the sense of "consisting essentially of", and its grammatical variations. Furthermore, the use of the term "a" or "an" are employed to describe elements and do not exclude the presence of more than one of such elements. Moreover, although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are generally only used to distinguish one element from another.

[0376] In some embodiments, numerical descriptions of components, quantities of attributes are used. It should be understood that such numerical descriptions used in the description of the embodiments are in some examples modified by the adjectives "approximately," "about," or "substantially." Unless otherwise indicated, "approximately," "about," or "substantially" means ±20% of the value stated. Accordingly, numerical parameters in the specification and claims are approximations, which can vary depending on the desired properties sought to be obtained by the particular embodiments. In some embodiments, numerical parameters are determined by the use of standard techniques. Although the numerical ranges and parameters setting forth the broad scope of the embodiments of the specification are approximations, unless otherwise indicated, these numerical values are to be understood as being modified in all instances by the term "about." In some embodiments, the numerical parameters are approximations that can vary depending on the desired properties sought to be obtained by the particular embodiments. In some embodiments, the numerical parameters are determined by the use of standard techniques.

[0377] Each patent, patent application, patent publication, and other material cited in this specification is hereby incorporated by reference in its entirety herein for the teachings relevant to the sentence and / or paragraph in which the reference is presented. Document histories, to the extent not inconsistent with the pertinent U.S. patent application file history, are also incorporated by reference herein. To the extent that material incorporated by reference contradicts or contradicts any portion of this specification, including definition, the portion of the material incorporated by reference prevails. Note, however, that in the event of inconsistencies between any such material and the present specification, including definitions, the present specification, including definitions, will control.

[0378] Finally, it should be understood that the embodiments described herein are merely exemplary of the principles of the present description. Other embodiments can be devised without departing from the scope of the present description. Accordingly, the embodiments described herein are not intended to limit the scope of the present description, but rather are intended to be exemplary thereof.

Claims

A method for detecting positioning abnormality of an automated mobile device, characterized by The method comprises: obtaining a first pose deviation of a first automatic mobile device relative to each of a plurality of first docking objects; obtaining a second pose deviation of each of one or more second automatic mobile devices relative to each of a plurality of second docking objects; determining a pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the first pose deviation of at least some of the plurality of first docking objects and the second pose deviation of at least some of the plurality of second docking objects; and determining whether a positioning device of the first automatic mobile device is abnormal based on the pose deviation similarity. The method of claim 1, wherein, The determining a pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the first pose deviation of at least some of the plurality of first docking objects and the second pose deviation of at least some of the plurality of second docking objects comprises: determining a total number of same docking objects in the plurality of first docking objects and the plurality of second docking objects; determining a target first pose deviation of the first automatic mobile device and a target second pose deviation of each of the second automatic mobile devices for determining the pose deviation similarity based on the total number, the target first pose deviation comprising a first pose deviation of the first automatic mobile device relative to a target first docking object in the plurality of first docking objects, and the target second pose deviation comprising a second pose deviation of each of the second automatic mobile devices relative to a target second docking object in the plurality of second docking objects; and determining the pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the target first pose deviation and the target second pose deviation. The method of claim 2, wherein, The determining a target first pose deviation of the first automatic mobile device and a target second pose deviation of each of the second automatic mobile devices for determining the pose deviation similarity comprises: in response to determining that the total number is less than a first threshold, determining the first pose deviation of the plurality of first docking objects as the target first pose deviation and determining the second pose deviation of the plurality of second docking objects as the target second pose deviation; in response to determining that the total number is greater than the first threshold, determining the first pose deviation of at least some of the same docking objects as the target first pose deviation and determining the second pose deviation of the at least some of the same docking objects as the target second pose deviation. The method of claim 3, wherein, The determining the first pose deviation of each of the at least some of the same docking objects as the target first pose deviation and determining the second pose deviation of each of the at least some of the same docking objects as the target second pose deviation in response to determining that the total number is greater than the first threshold further comprises: determining the first pose deviation of each of the at least some of the same docking objects as the target first pose deviation and determining the second pose deviation of each of the at least some of the same docking objects as the target second pose deviation based on a second threshold. In response to determining that the total number is greater than the first threshold and less than a second threshold, determining a first pose deviation of the same docking object as the target first pose deviation and a second pose deviation of the same docking object as the target second pose deviation; and, In response to determining that the total number is greater than the second threshold, determining a first pose deviation of a portion of the same docking object as the target first pose deviation and a second pose deviation of the portion of the same docking object as the target second pose deviation. The method of claim 4, wherein, The number of the portion of the same docking object is equal to the second threshold. The method of claim 2-5, wherein, The determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the target first pose deviation and the target second pose deviation includes: clustering the target first pose deviations to determine a first cluster center; clustering the target second pose deviations to determine a second cluster center; and determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first cluster center and the second cluster center. The method of claim 6, wherein, The total number of the same docking object is less than a first threshold. The method of claim 2-5, wherein, The determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first pose deviations corresponding to at least a portion of the plurality of first docking objects and the second pose deviations corresponding to at least a portion of the plurality of second docking objects includes: determining candidate pose deviation similarities based on the target first pose deviation and the target second pose deviation of each of the at least a portion of the same docking objects; and determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the candidate pose deviation similarities corresponding to the at least a portion of the same docking objects. The method of claim 8, wherein, The total number of the same docking object is greater than a first threshold. The method of claim 2-5, wherein, The determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first pose deviations corresponding to at least a portion of the plurality of first docking objects and the second pose deviations corresponding to at least a portion of the plurality of second docking objects includes: determining candidate pose deviation similarities based on the target first pose deviation and the target second pose deviation of each of the same docking objects; and determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the candidate pose deviation similarities corresponding to the same docking objects. The method of claim 10, wherein, The total number of the same docking object is greater than a first threshold and less than a second threshold. The method of claim 1, wherein The determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first pose deviations corresponding to at least a portion of the plurality of first docking objects and the second pose deviations corresponding to at least a portion of the plurality of second docking objects includes: determining a plurality of candidate pose deviation similarities based on the first pose deviations and the second pose deviations; and determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the plurality of candidate pose deviation similarities. The method of claim 1, wherein, The determining the pose deviation similarity between the first automated mobile device and each second automated mobile device based on the first pose deviations corresponding to at least a portion of the plurality of first docking objects and the second pose deviations corresponding to at least a portion of the plurality of second docking objects includes: determine a first representative pose deviation based on the first pose deviations; determine a second representative pose deviation based on the second pose deviations; and determine a pose deviation similarity between the first mobile device and each of the second mobile devices based on the first representative pose deviation and the second representative pose deviation. The method of claim 1-13, wherein, The determining whether the positioning device of the first mobile device is abnormal based on the pose deviation similarity comprises: determine a target pose similarity based on the pose deviation similarities between the first mobile device and the second mobile devices; and determine whether the positioning device of the first mobile device is abnormal based on the target pose similarity. A method for position compensation of an automated mobile device, characterized in that The method comprises: obtain a reference first positioning error of a target mobile device docking a target docking position, the reference first positioning error being caused by a positioning device of the target mobile device; obtain a reference second positioning error of the target mobile device docking the target docking position, the reference second positioning error being caused by a calibration of the target docking position; compensate the target docking position based on the reference first positioning error and the reference second positioning error to obtain a compensated target docking position; and control the mobile device to move to the compensated target docking position based on the compensated target docking position. The positioning compensation method of an automatic mobile device according to claim 15, characterized in that, The reference first positioning error and / or the reference second positioning error are pre-set and stored in a storage device, the storage device storing reference first positioning errors and / or reference second positioning errors of a plurality of mobile devices at any one of a plurality of docking positions. The positioning compensation method of an automated mobile device according to claim 15 or 16, characterized in that The reference first positioning error is determined based on the following operations: determine a first initial positioning error of the target mobile device at each of a plurality of docking positions; and determine the reference first positioning error based on the first initial positioning errors at the plurality of docking positions. The method of claim 17, wherein, The determining a first initial positioning error of the target mobile device at each of a plurality of docking positions comprises: after the target mobile device moves to a first actual arrival position related to each of the docking positions, obtain first perception data collected by the target mobile device at the first actual arrival position, the first perception data representing a first perceived position perceived by the target mobile device at the first actual arrival position; obtain a first error between the first actual arrival position and the docking position; determine a second error between the first perceived position and the docking position based on the first perception data; and determine the first initial positioning error of the target mobile device at each of the docking positions according to the first error and the second error. The method of claim 15-18, wherein The reference second positioning error is determined based on the following operations: adjust the target docking position based on the reference first positioning error to obtain an adjusted target docking position; acquire second perception data collected by the target automated mobile device at a second actual arrival position related to the target docking position, the second perception data representing a second perceived position perceived by the target automated mobile device at the second actual arrival position; acquire a third error between the second actual arrival position and the target docking position; determine a fourth error between the second perceived position and the target docking position based on the second perception data; and determine the reference second positioning error of the target automated mobile device at the target docking position according to the third error and the fourth error. The method of claim 18 or 19, wherein The first perception data or the second perception data is determined based on detection information acquired by the target automated mobile device scanning a marker. The method of claim 18 or 19, wherein The second error is determined based on the first perception data and reference position information of the target docking position in a reference map; and the fourth error is determined based on the second perception data and the reference position information of the target docking position in the reference map. The method of claim 15-21, wherein The method further comprises: determining whether the reference position information of the target docking position in the reference map and actual position information of the target docking position in a physical world satisfy a preset requirement; in response to determining that the reference position information of the target docking position in the reference map and the actual position information of the target docking position in the physical world do not satisfy the preset requirement, adjusting the reference position information of the target docking position in the reference map until the preset requirement is satisfied. The method of claim 22, wherein, Determining whether the reference position information of the target docking position in the reference map and actual position information of the target docking position in a physical world satisfy a preset requirement further comprises: acquiring third perception data collected by a plurality of automated mobile devices at the target docking position; determining third perceived positions of the plurality of automated mobile devices at the target docking position based on the third perception data; and determining whether the reference position information of the target docking position in the reference map and actual position information of the target docking position in a physical world satisfy a preset requirement based on whether a deviation between the third perceived positions and the reference position information of the target docking position in the reference map is within a preset range. The method of claim 15-23, wherein the method further comprises: The compensating the target docking position based on the reference first positioning error and the reference second positioning error to acquire a compensated target docking position comprises: acquiring a sum of the reference first positioning error and the reference second positioning error; determining a sum between the sum of the reference first positioning error and the reference second positioning error and the target docking position as the compensated target docking position. A positioning anomaly detection system for an autonomous mobile device, characterized by The positioning anomaly detection system comprises: an acquisition module, configured to acquire a first pose deviation of a first automated mobile device relative to each first docking object in a plurality of first docking objects; The acquisition module is further configured to acquire a second pose deviation of each of one or more second automatic mobile devices relative to each of a plurality of second docking objects; The determination module is configured to determine a pose deviation similarity between the first automatic mobile device and each of the second automatic mobile devices based on the first pose deviation and the second pose deviation; and The judgment module is configured to determine whether the positioning device of the first automatic mobile device is abnormal based on the pose deviation similarity. An apparatus for detecting a positioning abnormality of an automated mobile device, characterized by The positioning abnormality detection apparatus includes at least one processor and at least one memory; The at least one memory is configured to store computer instructions; The at least one processor is configured to execute at least part of the computer instructions to implement the method according to any one of claims 1-14. A computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, implement the positioning abnormality detection method of the automatic mobile device according to any one of claims 1-14. A positioning compensation system for an automated mobile device, characterized by The positioning compensation system includes: The error acquisition module is configured to acquire a reference first positioning error of a target automatic mobile device docking a target docking position, the reference first positioning error being caused by a positioning device of the target automatic mobile device; The error acquisition module is further configured to acquire a reference second positioning error of the target automatic mobile device docking the target docking position, the reference second positioning error being caused by a calibration of the target docking position; The compensation module is configured to compensate the target docking position based on the reference first positioning error and the reference second positioning error to acquire a compensated target docking position; and The control module is configured to control the automatic mobile device to move to the compensated target docking position based on the compensated target docking position. A positioning compensation device for an automated mobile device, characterized in that The positioning compensation apparatus includes at least one processor and at least one memory; The at least one memory is configured to store computer instructions; The at least one processor is configured to execute at least part of the computer instructions to implement the positioning compensation method of the automatic mobile device according to any one of claims 15-24. A computer readable storage medium, the storage medium storing computer instructions, when the computer instructions are executed by a processor, implement the positioning compensation method of the automatic mobile device according to any one of claims 15-24.

Images