Robot control method and device, cleaning equipment and computer program product

Abstract

The embodiment of the invention provides a robot control method, and relates to the technical field of robots, the method is applied to a robot provided with a line laser module, and when the robot is dragged in the process of executing a moving task, the robot is triggered to execute a rotating action; during execution of the rotation action, supplementary collection point cloud data are obtained through laser beam detection; and performing positioning recovery based on the complementary collection point cloud data so as to execute a mobile task. The robot is triggered to execute the rotation action during dragging, so that the line laser module can perform supplementary acquisition of point cloud data, positioning recovery can be completed, data instability caused by a small visual field of the line laser module is avoided, and the positioning accuracy is improved. And the task execution stability when the robot is dragged is improved.

Description

Technical Field

[0001] This specification relates to the field of robotics technology, specifically to robot control technology within the field of robotics technology, and more specifically to a robot control method, device, cleaning equipment, and computer program product. Background Technology

[0002] With the rapid development of service robots, industrial inspection robots and other fields, robots need to move autonomously and complete tasks in complex and dynamic environments, and accurate positioning and mapping are the core prerequisites for achieving autonomous movement.

[0003] In real-world applications, robots often face human interference, with human dragging being a typical form of interference. When a robot is dragged, its actual trajectory deviates significantly from the trajectory predicted by its internal positioning system. If the positioning and mapping system fails to recognize this anomaly in time and continues to update the map according to the normal movement state, errors such as map misalignment and overlap will occur, leading to subsequent positioning failures, path planning deviations, and even the robot's inability to complete its intended task.

[0004] Therefore, how to adjust control when a robot is being dragged by a person in order to continue the execution of the task has become a key problem that urgently needs to be solved in the field of robot control. Summary of the Invention

[0005] This specification provides a robot control method, device, computing equipment, and computer program product to improve the stability of task execution when a robot experiences a dragging event.

[0006] To achieve the above technical objectives, the embodiments of this specification provide the following technical solutions: Firstly, one embodiment of this specification provides a method for controlling a robot, the robot being a self-moving cleaning robot, the self-moving cleaning robot including a main body, a line laser module being provided on the front side of the main body along the direction of travel, the line laser module including a transmitting unit and a receiving unit, the line laser module being used to emit a first line laser beam emitted horizontally and a second line laser beam emitted obliquely downwards relative to the horizontal plane through the transmitting unit, the method comprising: When the robot is dragged during the execution of a movement task, it is triggered to perform a rotation action. The movement task is supported by the first line laser beam and the second line laser beam. During the rotation operation, a line laser beam is emitted into the target scene through the transmitting unit, and the first line laser beam reflected in the target scene along the horizontal direction by the receiving unit is received to obtain supplementary point cloud data. Based on the supplementary point cloud data, the robot's positioning is restored in order to perform the movement task.

[0007] Secondly, one embodiment of this specification provides a robot control device, comprising: The detection unit is used to trigger the robot to perform a rotation action when dragging occurs during the robot's movement task, and the movement task is supported by the first line laser beam and the second line laser beam. The detection unit is used to emit a line laser beam into the target scene through the transmitting unit during the rotation action, and to receive the first line laser beam reflected in the target scene along the horizontal direction through the receiving unit, so as to obtain supplementary point cloud data. The control unit is used to perform the robot's positioning recovery based on the supplementary point cloud data in order to perform the movement task.

[0008] Optionally, in one possible implementation, the detection unit is specifically used to acquire scene point cloud data corresponding to a preset time period through the line laser module and acquire inertial data corresponding to the preset time period through the robot's inertial detection module during the robot's mobile task execution. The detection unit is specifically used to compare the scene point cloud data with the memory map corresponding to the target scene to obtain a first abnormal parameter, and to obtain a second abnormal parameter by determining the fluctuation amplitude corresponding to the inertial data. The detection unit is specifically used to trigger the robot to perform a rotation action when the first abnormal parameter and the second abnormal parameter indicate that the robot is being dragged.

[0009] Optionally, in one possible implementation, the detection unit is specifically used to initialize the initial time parameters corresponding to the robot; The detection unit is specifically used to obtain scene time parameters corresponding to the target scene in response to the robot's movement task in the target scene; The detection unit is specifically configured to configure the initial time parameter based on the scene time parameter, so as to determine the preset time period according to the configured initial time parameter; The detection unit is specifically used to acquire scene point cloud data corresponding to the preset time period through the robot's line laser module, and to acquire inertial data corresponding to the preset time period through the robot's inertial detection module.

[0010] Optionally, in one possible implementation, the detection unit is specifically used to obtain the size parameters corresponding to the target scene and the reference size corresponding to the scene time parameters; The detection unit is specifically used to determine the area size of the scene where the robot is located based on the size parameters. The detection unit is specifically used to determine adjustment parameters based on the ratio of the area size to the reference size; The detection unit is specifically used to adjust the scene time parameter through the adjustment parameter, and configure the initial time parameter based on the adjusted scene time parameter, so as to determine the preset time period according to the configured initial time parameter.

[0011] Optionally, in one possible implementation, the detection unit is specifically used to compare the map at different acquisition times in the scene point cloud data with the memory map corresponding to the target scene to obtain the point cloud matching score corresponding to different acquisition times. The detection unit is specifically used to compare the point cloud matching score with the matching threshold to determine abnormal point cloud data. The detection unit is specifically used to determine the first abnormal parameter based on the proportion of the abnormal point cloud data to the scene point cloud data; The detection unit is specifically used to compare the detection data points in the inertial data with preset inertial parameters to determine the fluctuation amplitude. The detection unit is specifically used to compare the fluctuation amplitude with the fluctuation threshold to determine inertial anomaly data; The detection unit is specifically used to determine the second abnormal parameter based on the proportion of the inertial abnormal data to the total inertial data.

[0012] Optionally, in one possible implementation, the detection unit is specifically used to obtain the matching threshold corresponding to the target scene; The detection unit is specifically used to determine the scene complexity corresponding to the area where the robot is located based on the distribution of obstacle elements in the memory map. The detection unit is specifically used to adjust the matching threshold according to the scene complexity; The detection unit is specifically used to compare the point cloud matching score with the adjusted matching threshold to determine the abnormal point cloud data.

[0013] Optionally, in one possible implementation, the detection unit is specifically used to acquire the robot's historical inertial data in the target scene; The detection unit is specifically used to determine the fluctuation reference range indicated by the historical inertial data; The detection unit is specifically used to configure the fluctuation threshold according to the fluctuation reference range; The detection unit is specifically used to compare the fluctuation amplitude with the fluctuation threshold to determine the inertial anomaly data.

[0014] Optionally, in one possible implementation, the control unit is specifically used to issue the dragging state signal if, during the detection of the sensor parameters, both the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module meet preset conditions. The control unit is specifically used to control the robot to continuously update the first abnormal parameter and the second abnormal parameter based on the dragging status signal, and the scene map is prohibited from being updated during the dragging status signal output process; The control unit is specifically configured to trigger the robot to perform the rotation action in place if neither the updated first abnormal parameter nor the updated second abnormal parameter meets the preset conditions.

[0015] Optionally, in one possible implementation, the control unit is specifically used to issue a suspected dragging signal if, during the detection of the sensor parameters, both the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module meet preset conditions. The control unit is specifically used to trigger the robot to perform a rotation action based on the suspected dragging signal; The control unit is specifically used to control the third abnormal parameter collected by the line laser module and the fourth abnormal parameter collected by the inertial detection module during the execution of the rotation action. The control unit is specifically configured to issue the drag status signal if both the third abnormal parameter and the fourth abnormal parameter meet the preset conditions.

[0016] Thirdly, one embodiment of this specification also provides a cleaning device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the robot control method described above.

[0017] Fourthly, one embodiment of this specification also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the robot control method described above.

[0018] Fifthly, embodiments of this specification provide a computer program product or computer program, the computer program product including a computer program that can be stored in a computer-readable storage medium or in the cloud; the processor of the computer device reads the computer program, and when the processor executes the computer program, it implements the steps of the above-described robot control method.

[0019] As can be seen from the above technical solution, the robot control method provided in this specification triggers the robot to perform a rotation action when dragging occurs during the robot's movement task. This movement task is supported by a first line laser beam and a second line laser beam. During the rotation action, a line laser beam is emitted into the target scene through a transmitting unit, and the first line laser beam reflected in the target scene along the horizontal direction is received by a receiving unit to obtain supplementary point cloud data. Then, based on the supplementary point cloud data, the robot's positioning recovery is performed to execute the movement task. This achieves an intelligent drag event handling process. Because the robot's rotation action is triggered when dragging occurs, the line laser module can supplement point cloud data, thereby completing the positioning recovery. This avoids data instability caused by the small field of view of the line laser module and improves the stability of the robot's task execution when dragging occurs. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this specification. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0021] Figure 1 This is a schematic diagram illustrating the application environment of a robot control method provided in one embodiment of this specification.

[0022] Figure 2 This is a flowchart illustrating a robot control method provided as one embodiment of the present specification.

[0023] Figure 3 This is a schematic diagram of a cleaning device provided for one embodiment of this specification.

[0024] Figure 4 This is a schematic diagram of a line laser module provided for one embodiment of this specification.

[0025] Figure 5 This is a partially exploded view of a line laser module provided as one embodiment of this specification.

[0026] Figure 6 This is a schematic diagram of a line laser module for identifying light reflected from external objects, provided as one embodiment of this specification.

[0027] Figure 7 This is a schematic diagram of a robot control method provided as one embodiment of this specification.

[0028] Figure 8 This is a schematic diagram illustrating a scenario of another robot control method provided as one embodiment of this specification.

[0029] Figure 9 This is a schematic flowchart illustrating another robot control method provided as one embodiment of this specification.

[0030] Figure 10 This is a schematic diagram of the functional modules of a robot control device provided in one embodiment of this specification.

[0031] Figure 11 This is a schematic diagram of the structure of a computing device provided for one embodiment of this specification. Detailed Implementation

[0032] Unless otherwise defined, the technical or scientific terms used in the embodiments of this specification shall have the ordinary meaning understood by one of ordinary skill in the art to which this specification pertains. The terms "first," "second," and similar terms used in the embodiments of this specification do not indicate any order, quantity, or importance, but are merely used to avoid confusion of constituent elements.

[0033] Unless the context otherwise requires, throughout this specification, "a plurality of" means "at least two," and "including" is interpreted as open-ended or encompassing, that is, "including, but not limited to." In the description of this specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this specification. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example.

[0034] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0035] With the rapid development of service robots, industrial inspection robots and other fields, robots need to move autonomously and complete tasks in complex and dynamic environments, and accurate positioning and mapping are the core prerequisites for achieving autonomous movement.

[0036] In real-world applications, robots often face human interference, with human dragging being a typical form of interference. When a robot is dragged, its actual trajectory deviates significantly from the trajectory predicted by its internal positioning system. If the positioning and mapping system fails to recognize this anomaly in time and continues to update the map according to the normal movement state, errors such as map misalignment and overlap will occur, leading to subsequent positioning failures, path planning deviations, and even the robot's inability to complete its intended task.

[0037] Therefore, how to adjust control when a robot is being dragged by a person in order to continue the execution of the task has become a key problem that urgently needs to be solved in the field of robot control.

[0038] To address the aforementioned problems, this specification provides a robot control system, and the robot control method provided in this specification is applied to the robot's control system. This control system uses cooperating sensors, such as LiDAR and an inertial measurement unit (IMU), and monitors the corresponding data sliding window in real time. Then, it jointly determines whether the robot has been accidentally dragged by calculating environmental matching degree (a first anomaly parameter) and motion stability (a second anomaly parameter). Once dragging is confirmed, the system pauses routine tasks such as map building, initiates a positioning recovery procedure, and performs a rotational motion to re-acquire data, enabling the robot to re-determine its posture at the new location and resume the interrupted task.

[0039] Specifically, the robot's control system may include a system composed of... Figure 1The operating environment consists of a client, a server, and a robot. The client communicates with the server via a network. The client connects wirelessly to the robot. The robot communicates with the server via a network connection. The client can be an electronic device with network access capabilities. Specifically, for example, the client can be a desktop computer, tablet, laptop, smartphone, digital assistant, smart wearable device, shopping guide terminal, television, smart speaker, microphone, etc. Smart wearable devices include, but are not limited to, smart bracelets, smartwatches, smart glasses, smart helmets, smart necklaces, etc. Alternatively, the client can be software running on the electronic device. The server can be an electronic device with certain computing power. It can have a network communication module, processor, and memory, etc. Of course, the server can also refer to software running on the electronic device. The server can also be a distributed server, which can be a system with multiple processors, memory, network communication modules, etc., operating collaboratively. Alternatively, the server can be a cluster of several servers. Or, with the development of science and technology, the server can also be a new technical means capable of realizing the corresponding functions of the embodiments described in the specification. For example, it can be a new form of "server" based on quantum computing.

[0040] Specifically, based on the aforementioned control system, when the robot performs a task, a rotational motion is triggered when dragging occurs during the robot's movement. This movement is supported by a first and a second line laser beam. During the rotation, a line laser beam is emitted into the target scene via a transmitting unit, and the first line laser beam reflected horizontally in the target scene is received by a receiving unit to obtain supplementary point cloud data. Then, based on the supplementary point cloud data, the robot's positioning is restored to perform the movement task. This achieves intelligent drag event handling. Because the robot's rotation is triggered when dragging occurs, the line laser module can supplement point cloud data, thereby completing positioning restoration. This avoids data instability caused by the small field of view of the line laser module and improves the stability of the robot's task execution during dragging.

[0041] Based on the above concept, this specification provides a robot control method. The robot control method provided in this specification will be described exemplarily below with reference to the accompanying drawings.

[0042] To be applied Figure 1 Taking the robot in the example, some embodiments of this specification illustrate the control method of the robot, such as... Figure 2 As shown, Figure 2 A flowchart illustrating a robot control method according to one embodiment of this specification; the robot control method includes: 201. When dragging occurs during the robot's movement task, the robot is triggered to perform a rotation action. The movement task is supported by the first laser beam and the second laser beam.

[0043] In this embodiment, the robot can be a self-moving cleaning robot. For example, the cleaning equipment can be a robotic vacuum cleaner, an automatic sweeping vehicle, etc. Taking a robotic vacuum cleaner as an example, the cleaning equipment can be used to automatically or assistedly clean the floor. The cleaning equipment can reduce the physical exertion of manual cleaning through mechanization and automation technology, and can improve cleaning efficiency and convenience. The cleaning equipment in this embodiment can be applied to home or industrial environments, and is not limited thereto.

[0044] The cleaning equipment may include the equipment body and a line laser module. Regarding the installation method of the line laser module on the cleaning equipment, such as... Figure 3 As shown, Figure 3 This diagram illustrates a cleaning device according to one embodiment of the present specification; the diagram shows a line laser module disposed on the device body, specifically within the thickness direction. For example, the line laser module may be located at at least one of the front, rear, left, and right ends of the device body in the direction of travel.

[0045] Furthermore, the line laser module comprises a transmitting unit and a receiving unit. Specifically, the transmitting unit emits a first line laser beam emitted horizontally and a second line laser beam emitted at an angle downwards relative to the horizontal plane. Motion tasks can be performed with the support of these two line laser beams. The transmitting unit can emit the beam onto the surface of an external reflective object. The receiving unit can receive the beam reflected from the surface of the external reflective object to scan the surrounding environment, thereby achieving 3D mapping and obstacle detection. The emitted beam can include horizontal light and angled light. Horizontal light can be used for mapping functions of the cleaning equipment. An angled light can be used for obstacle avoidance functions of the cleaning equipment.

[0046] Specifically, regarding the implementation of the above functions, such as Figure 4 As shown, Figure 4This specification provides a schematic diagram of a line laser module according to one embodiment. The transmitting unit is a signal transmitting module used to transmit: a first line laser beam emitted horizontally and a second line laser beam emitted at an angle downwards relative to the horizontal plane. The first line laser beam corresponds to a first signal reflecting part, and the second line laser beam corresponds to a second signal reflecting part. That is, the signal transmitting module includes a first transmitting module and a second transmitting module. The first transmitting module emits the first line laser beam, and the second transmitting module emits the second line laser beam. The signal transmitting module and the reflector assembly are configured such that the first signal reflecting part can reflect the first line laser beam horizontally, and the second signal reflecting part can reflect the second line laser beam at an angle downwards relative to the horizontal plane, so that it can be received by the receiving unit (signal receiving module).

[0047] This application configures the signal transmitting module to emit a first linear laser beam emitted horizontally and a second linear laser beam tilted downward relative to the horizontal plane. The reflector assembly reflects the first linear laser beam horizontally, enabling distance detection in the horizontal direction and thus environmental detection. By reflecting the second linear laser beam tilted downward relative to the horizontal plane, it enables ground obstacle or drop detection. By configuring environmental detection and obstacle avoidance detection together, an automatic movement and cleaning process can be achieved.

[0048] The structure of the line laser module will be described below, as follows: Figure 5 As shown, Figure 5 This is a partially exploded structural diagram of a line laser module provided for one embodiment of this specification; the diagram shows that the line laser module 100 includes: a housing 110, a rotating mirror 150, and a mirror body 160.

[0049] The housing 110 may have mutually blocking emitting cavities 110a and receiving cavities 110b arranged along the thickness direction X of the housing 110. The housing 110 is provided with a light exit port 1114 corresponding to the emitting cavity 110a and a light inlet port 1112 corresponding to the receiving cavity 110b.

[0050] The rotating mirror 150 may be located within the housing 110. A portion of the rotating mirror 150 is located within the transmitting cavity 110a. Another portion of the rotating mirror 150 is located within the receiving cavity 110b.

[0051] A portion of the mirror body 160 may be located at the light exit port 1114. Another portion of the mirror body 160 may be located at the light inlet port 1112, and the portion of the mirror body 160 located at the light exit port 1114 includes a tilting portion. The tilting portion can at least be used to reflect a portion of the horizontal light projected onto the tilting portion into the housing 110 and avoid the rotating mirror 150.

[0052] It should be noted that the mirror body 160 can be transparent. The mirror body 160 can cover the light inlet 1112 and the light outlet 1114. The emitted light can pass through the mirror body 160 through the light outlet 1114 and be emitted towards the external reflector. The light reflected by the external reflector passes through. The light reflected by the external reflector can pass through the light inlet 1112 through the mirror body 160 and enter the receiving cavity 110b.

[0053] In this embodiment, the horizontal light M from the emission cavity 110a is projected towards the mirror body 160 under the action of the rotating mirror 150. Part of the horizontal light M can pass through the mirror body 160 and be projected onto external reflective objects. Part of the horizontal light M is reflected on the mirror body 160 to form reflected light N. By providing an inclined portion on the mirror body 160, the reflected light N generated by the horizontal light M on the mirror body 160 can avoid the direction of the rotating mirror 150. In other words, the reflected light N generated by the horizontal light M is less likely to be projected back onto the rotating mirror 150. Therefore, the possibility of the reflected light from the horizontal light being projected a second time onto the rotating mirror 150 and then onto other positions via the mirror body 160 can be reduced, which could lead to misjudging the position of external reflective objects as the position of false point clouds, resulting in misjudgment of the position of external reflective objects. This could reduce the detection accuracy of the cleaning equipment and affect the user experience.

[0054] In this embodiment, by providing an inclined portion, the reflected light N formed on the inclined portion can avoid the rotating mirror 150. The reflected light N can be deflected away from the rotating mirror 150, making it less likely for the reflected light N to be projected onto the rotating mirror 150 again. This reduces the possibility of the reflected light N being projected onto the rotating mirror 150 a second time, leading to misjudgment of the position of external reflective objects, and helps to ensure the detection accuracy of the cleaning equipment.

[0055] In one possible scenario, because mirror 160 can be transparent, horizontal light M is projected a second time onto rotating mirror 150. Rotating mirror 150 causes the reflected light N to be emitted in other directions, potentially forming a false point cloud. Therefore, the location of external reflective objects is misjudged as the location of the false point cloud, leading to misidentification of external reflective objects and reduced detection accuracy of the cleaning equipment.

[0056] Therefore, in this embodiment, the mirror body 160 is designed as an inclined structure that tilts outward toward the housing 110, such as... Figure 6 As shown, Figure 6This specification provides a schematic diagram of a line laser module for identifying external reflective objects according to one embodiment. By setting the first region 161a as an inclined structure tilted outward from the housing 110, the reflected light N formed by the horizontal light M projected onto the first region 161a can be deflected away from the rotating mirror 150 under the action of the inclined structure. This makes it less likely that the reflected light N will be projected onto the rotating mirror 150 again, thereby reducing the possibility of the reflected light N being projected onto the rotating mirror 150 a second time, resulting in false point clouds and misjudging the position of external reflective objects. This helps to ensure the detection accuracy of the cleaning equipment.

[0057] Based on the aforementioned line laser module, sensor parameter detection can be performed, i.e., the drag-and-drop judgment process.

[0058] In one possible scenario, the detection process of sensor parameters can be based on a line laser module and an inertial detection module. First, during the robot's movement task, the line laser module acquires scene point cloud data corresponding to a preset time period, and the robot's inertial detection module acquires inertial data corresponding to the preset time period. Then, the scene point cloud data is compared with the memory map corresponding to the target scene to obtain a first abnormal parameter, and a second abnormal parameter is obtained by determining the fluctuation amplitude corresponding to the inertial data. When the first and second abnormal parameters indicate that the robot is dragging, the robot is triggered to perform a rotation action.

[0059] Among them, the robot is a mobile robot with specific functions (such as sweeping and carrying), and its movement method includes but is not limited to rollers, tracks, etc., depending on the actual scenario; correspondingly, the target scenario is the place where the robot performs the task, such as a house where the robot performs the sweeping task, or a warehouse where the robot performs the carrying task.

[0060] Specifically, the robot is equipped with a line laser module for environmental perception and an inertial detection module for inertial detection. The line laser module is a new type of lidar that can scan the environmental point cloud 120° in front of the robot using a laser transceiver and a rotating mirror. This radar module is installed directly in front of the robot, which can reduce the robot's height, improve the robot's ability to enter low-ceilinged spaces, expand the robot's operating space, and has a cost advantage compared to traditional lidar. However, since the field of view of the point cloud scanned by the line laser is only 120°, the confidence of the point cloud matching with the map is lower than that of 360° lidar. Therefore, the algorithm-based localization recovery scheme of line lidar will not use point clouds with low matching scores for plotting, and it is impossible to create a sub-map for detecting localization anomalies. Furthermore, the reliability of 120° laser point clouds is low when performing global relocalization and cannot be directly used for localization recovery. Therefore, this embodiment is designed with targeted improvements based on these issues.

[0061] In addition, the inertial detection module can be an inertial measurement unit (IMU), which is a sensor used to measure the robot's angular velocity and acceleration, and can output the robot's motion state data in real time to assist in positioning calculations.

[0062] It is understandable that the above sensors collect data through a preset time period. Within the preset time period, data can be understood as being collected through a sliding window. Different sliding windows are configured for different sensors. That is, data from the corresponding sensors is collected through a dual window with an association relationship (the data collected by the line laser module within the preset time period corresponds to the first sliding window, and the data collected by the inertial detection module within the preset time period corresponds to the second sliding window, hereinafter collectively referred to as sliding windows). This can match the data characteristics of the sensors. That is, each window is a data queue of a fixed size, storing data points from the most recent period.

[0063] In one possible scenario, the initialization and configuration process of the sliding window can be described. The sliding window can use a fixed window size or be configured according to the specific "target scenario." That is, when the robot enters different work areas or performs different tasks, the duration or amount of data used for anomaly statistics can be dynamically adjusted, making the detection mechanism more adaptable to different scenarios.

[0064] Specifically, for configuring the preset time period, i.e. the sliding window configuration process, we can first initialize the robot's corresponding initial time parameters (equivalent to preset window parameters); then, in response to the robot's movement task in the target scene, we can obtain the scene time parameters corresponding to the target scene (equivalent to sliding window parameters); and configure the initial time parameters based on the scene time parameters to determine the preset time period according to the configured initial time parameters; then, we can collect data through the preset time period, i.e., obtain the scene point cloud data corresponding to the first sliding window through the robot's line laser module, and obtain the inertial data corresponding to the second sliding window through the robot's inertial detection module.

[0065] Take a warehouse robot as an example. When the robot starts up or enters a new area, the system first initializes initial time parameters (such as setting the default window size to store data from the last 10 seconds). Then, the system reads the preset configuration for this warehouse area (target scene). This configuration might suggest using a shorter window (e.g., 5 seconds) in narrow aisles for faster response, and a longer window (e.g., 15 seconds) in open shelving areas to filter out occasional interference. The robot system then configures itself based on the scene time parameters, determining the final sizes of the laser data window (first sliding window) and the IMU data window (second sliding window), and then begins collecting and monitoring data using these two configured windows.

[0066] Take a robotic vacuum cleaner as an example. When the robotic vacuum cleaner is working in different rooms, the system dynamically adjusts its detection sensitivity. For instance, before entering a narrow area, the system reads the scene time parameters for that area (it is recommended to use a short window for faster response) and adjusts the default 10-second detection window (initial time parameter) to 5 seconds. Therefore, in confined spaces, once the robot is moved, it can detect the anomaly in a shorter time (5-second window), improving detection accuracy.

[0067] In another possible scenario, to improve the adaptability of the sliding window configuration, a mechanism can be introduced to dynamically adjust the window parameters based on the size of the area where the robot is located. That is, in confined spaces where environmental features change rapidly, more sensitive detection (smaller window) is needed; in open spaces, more robust detection (larger window) is possible.

[0068] Therefore, for the configuration process of the first and second sliding windows, the size parameters corresponding to the target scene and the reference size corresponding to the scene time parameters can be obtained first; then the area size of the scene position where the robot is located can be determined based on the size parameters; and the adjustment parameters can be determined according to the ratio of the area size to the reference size; then the scene time parameters can be adjusted by adjusting the adjustment parameters, and the initial time parameters can be configured based on the adjusted scene time parameters, so as to determine the preset time period according to the configured initial time parameters.

[0069] Take a warehouse robot as an example. When the robot moves through the warehouse, it senses the size of its environment in real time. When it enters a narrow corridor (a very small area) that is only slightly wider than itself, the system calculates an adjustment parameter based on the ratio of this size to a reference size (such as the size of an open hall). (For example, if the ratio is less than 0.5, the adjustment parameter is 0.7). The system uses this parameter to reduce the preset window size (e.g., adjusting the preset 10-second window to 7 seconds), making anomaly detection more sensitive and faster in narrow areas where caution is required. Conversely, when entering an open area, the window will return to its original size or become larger to enhance its anti-interference capabilities.

[0070] Take a robot vacuum cleaner as an example. See Figure 7 The scene shown, Figure 7 This is a schematic diagram illustrating a robot control method according to one embodiment of this specification. When the robot vacuum cleaner moves from a spacious area 1 into a narrow area 2, the system calculates the width (area size) of area 2 in real time. It is found that the width of area 2 is less than 1 meter (while the reference size of area 1 is 5 meters), a ratio of approximately 0.2. Based on this ratio, the system dynamically adjusts the detection window size from the default 10 seconds to 2 seconds (adjustment parameter 0.2), making anomaly detection extremely sensitive in smaller areas (such as corridors), allowing for rapid detection even of slight movements.

[0071] After acquiring sensor data through the dual sliding window configuration, the anomaly detection process is carried out. Specifically, the point cloud data can be compared with the memory map corresponding to the target scene to obtain the first anomaly parameter, and the second anomaly parameter can be obtained by determining the fluctuation amplitude corresponding to the inertial data.

[0072] In this embodiment, regarding the configuration of the preset time period, i.e., the configuration of the first and second abnormal parameters, since the data obtained through the sliding window is a time-based data sequence, both can use the proportion of abnormal data points as the output, or they can use the number of abnormal data points or extreme values, etc. The calculation of the proportion of abnormal data points is as follows: the first abnormal parameter is based on the proportion of point cloud matching scores below a threshold; the second abnormal parameter is based on the proportion of inertial data (such as Z-axis acceleration) fluctuations exceeding a threshold. Using the proportion statistics method can effectively smooth instantaneous noise and improve the stability of the judgment.

[0073] Specifically, for the process of configuring abnormal parameters based on the proportion of abnormal data points, we can first compare the map at different collection times in the scene point cloud data with the memory map corresponding to the target scene to obtain the point cloud matching score corresponding to different collection times; then compare the point cloud matching score with the matching threshold to identify abnormal point cloud data; and determine the first abnormal parameter based on the proportion of abnormal point cloud data to the scene point cloud data.

[0074] For anomalies in inertial data, the detected data points in the inertial data can be compared with preset inertial parameters to determine the fluctuation amplitude; the fluctuation amplitude can be compared with the fluctuation threshold to determine the inertial anomalous data; and then, based on the proportion of the inertial anomalous data to the total inertial data, a second anomalous parameter can be determined.

[0075] For example, for LiDAR, the currently scanned shelf point cloud can be matched with a memory map to calculate a point cloud matching score (e.g., 0.85 out of 1.0). If the score is lower than the matching threshold (e.g., 0.6), the data at that moment is considered a point cloud anomaly. If 7 out of the last 10 scans (windows) fail to match, the first anomaly parameter (point cloud anomaly percentage) is 70%. For IMU, the system continuously reads the Z-axis acceleration, which should normally fluctuate slightly around 9.8 m / s² (gravitational acceleration). If the acceleration value deviates from the gravitational acceleration by more than the fluctuation threshold (e.g., 0.5 m / s²) at a certain moment, the data at that moment is considered inertial anomaly data. If 80 out of the last 100 IMU samples show abnormal fluctuations, the second anomaly parameter (inertial anomaly percentage) is 80%.

[0076] In one possible scenario, considering the diversity of robot operating environments, an adaptive adjustment strategy can be introduced for the point cloud matching threshold. That is, the threshold is not fixed but dynamically adjusted based on the scene complexity at the robot's current location (determined by the distribution of obstacles in the map). In feature-rich and structurally complex regions, the threshold can be appropriately relaxed; in feature-sparse and highly repetitive regions, the threshold needs to be tightened to reduce misjudgments.

[0077] Therefore, the configuration process for point cloud anomaly data can be achieved by obtaining the matching threshold corresponding to the target scene; then, based on the distribution of obstacle elements in the memory map, the scene complexity corresponding to the area where the robot is located can be determined; and the matching threshold can be adjusted according to the scene complexity; finally, the point cloud matching score can be compared with the adjusted matching threshold to identify point cloud anomaly data.

[0078] Take warehouse robots as an example. In densely packed shelving areas of a warehouse, laser scanning can obtain a large number of clear and unique shelving contour features, resulting in high matching reliability. In this case, the system can appropriately increase the matching threshold (e.g., from 0.6 to 0.7), and only highly accurate matches are considered normal. This helps maintain a high standard of positioning accuracy even in complex environments. Conversely, when the robot passes through an empty, smooth wall, the laser point cloud features are very simple, and the matching score may be low and unstable. In this case, the system will identify that the scene complexity in this area is low and correspondingly lower the matching threshold (e.g., from 0.6 to 0.5) to avoid frequent false alarms of "point cloud matching anomalies" due to the scarcity of features in the environment itself.

[0079] Take a robot vacuum cleaner as an example. Figure 8 The scene shown, Figure 8 This diagram illustrates a scenario for another robot control method provided in one embodiment of this specification. In a living room with dense furniture, laser scanning can detect multiple obstacles (sofa legs, table corners, wall corners, etc.), increasing the likelihood of being dragged. The system recognizes the high scene complexity in this area and raises the matching threshold from the default 0.6 to 0.7, only considering a very high matching degree as normal. Conversely, on an open balcony with sparse features, the system recognizes low scene complexity and lowers the matching threshold to 0.4 to avoid frequent false alarms due to the limited features of the environment itself.

[0080] Furthermore, for configuring the fluctuation threshold, historical inertial data of the robot during normal movement in the target scene can be used to learn and determine a reasonable reference range for "normal fluctuations." Configuring the threshold based on this historical statistical information makes anomaly detection more consistent with the robot's actual motion characteristics in that scene, avoiding false alarms or missed alarms caused by inappropriate preset thresholds. Specifically, the process involves first acquiring the robot's historical inertial data in the target scene; then determining the fluctuation reference range indicated by the historical inertial data; configuring the fluctuation threshold based on the fluctuation reference range; and finally comparing the fluctuation amplitude with the fluctuation threshold to identify inertial anomalies.

[0081] Take a warehouse robot as an example. The degree of bumpiness varies depending on the surface. When the robot enters a warehouse area covered with anti-slip rubber mats, it first records a segment of IMU data during normal movement as historical inertial data. Analysis reveals that due to the slight elasticity of the ground, the Z-axis acceleration fluctuates within a reference range of ±0.4 m / s² during normal movement, which is larger than on smooth concrete (±0.2 m / s²). Therefore, the system configures the fluctuation threshold to 0.5 m / s² based on this learned reference range, instead of the globally default 0.3 m / s². Thus, only abnormally severe bumps (exceeding 0.5 m / s²) are considered abnormal, adapting to the actual conditions of the special surface.

[0082] Take a robot vacuum cleaner as an example. When working in a carpeted bedroom, the softness of the carpet causes significant vibrations during normal movement. The system first records a segment of IMU data during normal cleaning as historical inertial data. Analysis reveals that the normal fluctuation range of Z-axis acceleration is ±0.4 m / s² (larger than on hard surfaces). Therefore, the system adjusts the fluctuation threshold from the default 0.3 m / s² to 0.6 m / s², only classifying abnormal vibrations (exceeding 0.6 m / s²) as abnormal, thus preventing large vibrations during normal carpet cleaning from being mistaken for dragging.

[0083] In this embodiment, the output of the dragging status signal after determining that the robot has dragged can be based on a parameter comparison between the first abnormal parameter and the second abnormal parameter. If both the first abnormal parameter and the second abnormal parameter meet the preset conditions, the dragging status signal is output, which is used to indicate that the robot has dragged.

[0084] The preset condition can be a threshold for the proportion of abnormal data, such as 0.8. Both the first and second abnormal parameters must meet the preset condition, that is, the numbers of both the first and second abnormal parameters must be greater than 0.8.

[0085] Therefore, when the matching degree between the environmental scene point cloud data collected by the robot in real time through the line laser module and the preset memory map is consistently low, and at the same time the inertial detection module (such as IMU) detects that the robot's Z-axis acceleration is constantly fluctuating violently due to non-autonomous motion, the system determines that these two anomalies occur simultaneously within the time window and both exceed the normal range. At this time, a dragging status signal will be triggered and output. This signal indicates that the robot is currently being dragged or moved by an external force, thereby activating the corresponding protection and recovery mechanism.

[0086] 202. During the rotation operation, a line laser beam is emitted to the target scene through the transmitting unit, and the first line laser beam reflected in the target scene along the horizontal direction is received through the receiving unit to obtain supplementary point cloud data.

[0087] In this embodiment, referring to the description of the line laser module in step 201, since the field of view of the point cloud of the line laser scanning is only 120°, the confidence of the point cloud matching with the map is lower than that of the 360° LiDAR. Therefore, the algorithm-based positioning and recovery scheme of the line LiDAR will not use the point cloud with the lower matching score for plotting, and thus it is impossible to create a sub-map for detecting positioning anomalies. In addition, the reliability of the 120° laser point cloud is low when performing global relocalization and cannot be directly used for positioning and recovery.

[0088] Understandably, in environmental detection, the emitted beam can include horizontal and oblique beams. Horizontal beams can be used to achieve mapping functions for cleaning equipment. Oblique beams can be used to achieve obstacle avoidance functions for cleaning equipment. However, during rotation, the operation is performed in place; therefore, only the beam is emitted to the target scene through the transmitting unit, and the beam reflected from the horizontal beam in the target scene is received by the receiving unit to obtain supplementary point cloud data.

[0089] In one possible scenario, the rotation action can be performed after determining that the robot is not being dragged. That is, if the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module both meet the preset conditions during the sensor parameter detection process, a dragging status signal is issued. Then, based on the dragging status signal, the robot is controlled to continuously update the first and second abnormal parameters. During the dragging status signal output, the scene map is prohibited from being updated. If the updated first and second abnormal parameters do not meet the preset conditions, the robot is triggered to perform a rotation action in place.

[0090] Understandably, once the robot confirms the dragging has occurred and outputs a dragging status signal, its primary task is to prevent updates to the scene map, thus avoiding the contamination of the already established high-precision map by erroneous perception data during the dragging process. After the robot is dragged to the new location and placed down, the laser matching and IMU readings gradually return to normal. When the system detects that the updated dual abnormal parameters (such as the point cloud abnormality rate dropping to 20% and the inertial abnormality rate dropping to 10%) do not meet the preset abnormality conditions, it determines that the dragging has ended. Immediately, the robot is triggered to slowly rotate 1 revolution in place, using the LiDAR to collect point cloud data from multiple angles, obtaining a 360-degree environmental snapshot. Next, the algorithm uses the last known correct position before the dragging as the center to crop a sufficiently large local map, and then matches the newly collected omnidirectional point cloud with this local map to quickly complete the local positioning recovery operation (i.e., repositioning). After successful repositioning, the system resumes map updates, and the robot can continue its movement and operation from the correct new position.

[0091] In addition, for the output of the drag status signal, a suspected judgment can be performed. If, during the detection of sensor parameters, the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module both meet the preset conditions, a suspected drag signal is issued. Then, based on the suspected drag signal, the robot is triggered to perform a rotation action. During the rotation action, the third abnormal parameter collected by the line laser module and the fourth abnormal parameter collected by the inertial detection module are controlled. If both the third and fourth abnormal parameters meet the preset conditions, a drag status signal is issued.

[0092] The third and fourth abnormal parameters are the data collected during the rotation operation after the first and second abnormal parameters have met the preset conditions.

[0093] As can be seen, in the process of suspected drag detection, when the proportion of abnormal data in both sliding windows simultaneously exceeds the threshold, the robot is triggered to rotate and collect additional point cloud data from multiple directions. The point cloud matching sliding window data is then reassessed. If the proportion of abnormal matching data is still higher than the threshold, the process enters the confirmation stage; otherwise, it returns to normal operation. This improves the accuracy of drag detection.

[0094] 203. Based on the supplementary point cloud data, perform the robot's positioning recovery to execute the mobile task.

[0095] In this embodiment, the positioning recovery operation refers to the recovery behavior of the dragging behavior, which ensures that the robot can continue to perform the task.

[0096] The following explanation, using the above embodiments, focuses on a scenario where a robot is equipped with a 360° LiDAR. This is because the SLAM algorithm based on 360° LiDAR handles localization anomalies by: identifying the anomaly through the matching deviation between the subgraph and the map; then restoring the localization through global relocalization; and finally reverting the map to the state where the matching was successful. This method has some limitations: firstly, subgraph matching and global relocalization consume significant computational resources; secondly, simply identifying localization anomalies through subgraph matching deviation cannot effectively detect human dragging, and may result in false detections due to dynamic environmental changes; furthermore, this detection method has a slow response time, and anomaly detection typically only becomes effective when the deviation accumulates to a certain level. During the period between human dragging of the robot and successful detection of the localization anomaly and the localization recovery and map reversion, the robot may still operate according to an incorrect map, posing a risk of further anomalies.

[0097] In addition to the limitations mentioned above, the data provided by line LiDAR is also not suitable for this traditional method of location anomaly detection and location recovery. Since the field of view of the point cloud of line LiDAR scanning is only 120°, the confidence of the point cloud matching with the map is lower than that of LDS radar. Therefore, the SLAM scheme of line LiDAR will not use the point cloud with the lower matching score for plotting, and it is impossible to create a sub-map for detecting location anomalies. Furthermore, the reliability of 120° LiDAR point cloud is low when performing global relocalization and cannot be directly used for location recovery.

[0098] To address the shortcomings of current methods, and taking advantage of the characteristics of line lidar, the method proposed in this invention can accurately detect and quickly respond to situations where a machine is being dragged by a person in an environment with a memory map. After the dragging is detected, the dragged machine can re-collect point clouds by rotating in place, thus compensating for the lack of data from the line lidar. Subsequently, it can restore its location through local relocation and continue to execute the current task. Moreover, the computational power consumption for drag detection and location restoration is lower than that of traditional methods.

[0099] For scenarios where the line laser module is a line LiDAR, the robot can perform tasks such as... Figure 9 Steps, Figure 9 A schematic flowchart illustrating another robot control method provided in one embodiment of this specification; the figure shows the following steps: 901. Normal operation phase.

[0100] During this phase, the robot moves normally and can initialize the sliding window for later use.

[0101] 902. Anomaly Detection Phase.

[0102] This stage can be performed using an anomaly detection module. After initializing the dual sliding windows, the LiDAR collects point cloud data and matches it with the memory map, while the IMU collects acceleration data. The dual sliding windows are updated using the point cloud matching score and z-axis acceleration data. When the point cloud matching score is lower than a threshold, it is considered an anomaly. When the robot is operating normally on the ground, the z-axis acceleration is usually stable at the gravitational acceleration. However, when the robot is being dragged, the z-axis acceleration will fluctuate. If the fluctuation amplitude is greater than a threshold, the acceleration data is considered abnormal. The proportion of matching anomalies and acceleration anomalies in the dual sliding windows is calculated separately. If the proportion of matching anomalies or acceleration anomalies is lower than a threshold, the robot is in normal operating mode, and the map can be updated.

[0103] 903. The dragging stage.

[0104] When the robot moves normally to point A, it is manually dragged to point B. During this period, if the percentage of abnormal data in both sliding windows simultaneously exceeds the threshold, the dragging phase is confirmed. The map control module immediately stops map updates, records the current map status, and continuously monitors the two sliding windows until the dragging ends.

[0105] 904. Location recovery phase.

[0106] This stage can be executed through the positioning recovery module. That is, after the dragging ends, the robot enters the positioning recovery stage at point B. At this time, the proportion of point cloud matching anomalies is lower than the threshold and the proportion of acceleration anomalies is lower than the threshold. The end of dragging is detected, triggering the robot to rotate in place through the motion control module. The line LiDAR re-collects point cloud data; the local map is cropped, local relocalization is performed, the robot pose is corrected, and the robot position is restored from the dragging start point A to the dragging end point B.

[0107] After the location is restored, the map update is resumed through the map control module, and mapping continues from the accurate location; the robot starts from point B and runs normally, and the anomaly detection module re-enters the double sliding window statistical loop.

[0108] As can be seen, the above example adapts to the differences in characteristics between LiDAR and IMU, adopts a "dual sliding window management" mechanism, and combines it with the robot's movement strategy. Through the collaborative work of four core functional modules—anomaly detection module, localization recovery module, map control module, and motion control module—it accurately distinguishes between human dragging and normal motion anomalies, solving the localization and mapping problems caused by human dragging, as shown in Table 1. Table 1 Functional Module Composition It is understood that the above module distribution is only an example, and the specific implementation may be a combination of various hardware containing the above module functions or the writing of algorithms, which is not limited here.

[0109] As can be seen, by adapting to sensor characteristics through dual sliding windows, false positives and false negatives are reduced, resulting in high accuracy in drag detection. Furthermore, the use of local matching combined with action-based data acquisition significantly improves efficiency compared to existing global relocation methods, leading to high location recovery efficiency. By disabling updates during dragging and resuming subsequent writing, the risk of map errors is reduced, resulting in strong map stability.

[0110] In summary, this embodiment triggers a rotational action when dragging occurs during a robot's movement task. This movement task is supported by a first and a second line laser beam. During the rotation, a line laser beam is emitted into the target scene via a transmitting unit, and the first line laser beam reflected horizontally in the target scene is received by a receiving unit to obtain supplementary point cloud data. Based on this supplementary point cloud data, the robot's positioning is restored to perform the movement task. This achieves intelligent drag event handling. Because the robot's rotation is triggered during dragging, the line laser module can supplement point cloud data, thus enabling positioning restoration. This avoids data instability caused by the limited field of view of the line laser module and improves the stability of the robot's task execution during dragging.

[0111] It should be noted that the various embodiments described in this specification emphasize the parts that differ from other embodiments, and the embodiments can be explained by comparison with each other. Any combination of the various embodiments described in this specification based on general technical knowledge is covered within the scope of this specification.

[0112] In one exemplary embodiment of this specification, a robot control device 1000 is also provided, such as... Figure 10 As shown, Figure 10 This specification provides a functional block diagram of a robot control device 1000 according to one embodiment of the present invention. The control device 1000 includes: The detection unit 1001 is used to trigger the robot to perform a rotation action when dragging occurs during the robot's movement task, wherein the movement task is supported by the first line laser beam and the second line laser beam. The detection unit 1001 is used to emit a line laser beam into the target scene through the transmitting unit during the rotation action, and to receive the first line laser beam reflected in the target scene along the horizontal direction through the receiving unit, so as to obtain supplementary point cloud data. The control unit 1002 is used to perform the robot's positioning recovery based on the supplementary point cloud data in order to perform the movement task.

[0113] Optionally, in one possible implementation, the detection unit 1001 is specifically used to acquire scene point cloud data corresponding to a preset time period through the line laser module and acquire inertial data corresponding to the preset time period through the robot's inertial detection module during the robot's mobile task execution. The detection unit 1001 is specifically used to compare the scene point cloud data with the memory map corresponding to the target scene to obtain a first abnormal parameter, and to obtain a second abnormal parameter by determining the fluctuation amplitude corresponding to the inertial data. The detection unit 1001 is specifically used to trigger the robot to perform a rotation action when the first abnormal parameter and the second abnormal parameter indicate that the robot is being dragged.

[0114] Optionally, in one possible implementation, the detection unit 1001 is specifically used to initialize the initial time parameters corresponding to the robot; The detection unit 1001 is specifically used to obtain scene time parameters corresponding to the target scene in response to the robot's movement task in the target scene; The detection unit 1001 is specifically used to configure the initial time parameter based on the scene time parameter, so as to determine the preset time period according to the configured initial time parameter; The detection unit 1001 is specifically used to acquire scene point cloud data corresponding to the preset time period through the robot's line laser module, and to acquire inertial data corresponding to the preset time period through the robot's inertial detection module.

[0115] Optionally, in one possible implementation, the detection unit 1001 is specifically used to obtain the size parameters corresponding to the target scene and the reference size corresponding to the scene time parameters; The detection unit 1001 is specifically used to determine the area size of the scene where the robot is located based on the size parameters. The detection unit 1001 is specifically used to determine adjustment parameters based on the ratio of the area size to the reference size; The detection unit 1001 is specifically used to adjust the scene time parameter through the adjustment parameter, and configure the initial time parameter based on the adjusted scene time parameter, so as to determine the preset time period according to the configured initial time parameter.

[0116] Optionally, in one possible implementation, the detection unit 1001 is specifically used to compare the map at different acquisition times in the scene point cloud data with the memory map corresponding to the target scene to obtain the point cloud matching score corresponding to different acquisition times. The detection unit 1001 is specifically used to compare the point cloud matching score with the matching threshold to determine abnormal point cloud data. The detection unit 1001 is specifically used to determine the first abnormal parameter based on the proportion of the abnormal point cloud data to the scene point cloud data. The detection unit 1001 is specifically used to compare the detection data points in the inertial data with preset inertial parameters to determine the fluctuation amplitude. The detection unit 1001 is specifically used to compare the fluctuation amplitude with the fluctuation threshold to determine the inertial anomaly data; The detection unit 1001 is specifically used to determine the second abnormal parameter based on the proportion of the inertial abnormal data to the total inertial data.

[0117] Optionally, in one possible implementation, the detection unit 1001 is specifically used to obtain the matching threshold corresponding to the target scene; The detection unit 1001 is specifically used to determine the scene complexity corresponding to the area where the robot is located based on the distribution of obstacle elements in the memory map. The detection unit 1001 is specifically used to adjust the matching threshold according to the scene complexity; The detection unit 1001 is specifically used to compare the point cloud matching score with the adjusted matching threshold to determine the abnormal point cloud data.

[0118] Optionally, in one possible implementation, the detection unit 1001 is specifically used to acquire the robot's historical inertial data in the target scene; The detection unit 1001 is specifically used to determine the fluctuation reference range indicated by the historical inertial data; The detection unit 1001 is specifically used to configure the fluctuation threshold according to the fluctuation reference range; The detection unit 1001 is specifically used to compare the fluctuation amplitude with the fluctuation threshold to determine the inertial anomaly data.

[0119] Optionally, in one possible implementation, the control unit 1002 is specifically used to issue the dragging state signal if, during the detection of the sensor parameters, both the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module meet preset conditions. The control unit 1002 is specifically used to control the robot to continuously update the first abnormal parameter and the second abnormal parameter based on the drag state signal, and the scene map is prohibited from being updated during the drag state signal output process; The control unit 1002 is specifically used to trigger the robot to perform the rotation action in place if neither the updated first abnormal parameter nor the updated second abnormal parameter meets the preset conditions.

[0120] Optionally, in one possible implementation, the control unit 1002 is specifically used to issue a suspected dragging signal if, during the detection of the sensor parameters, both the first abnormal parameter collected by the line laser module and the second abnormal parameter collected by the inertial detection module meet preset conditions. The control unit 1002 is specifically used to trigger the robot to perform a rotation action based on the suspected dragging signal; The control unit 1002 is specifically used to control the third abnormal parameter collected by the line laser module and the fourth abnormal parameter collected by the inertial detection module during the execution of the rotation action. The control unit 1002 is specifically used to issue the drag status signal if both the third abnormal parameter and the fourth abnormal parameter meet the preset conditions.

[0121] Specifically, the processing unit and control unit in this embodiment can correspond to physical components. For example, the processing unit can be a processing module such as a CPU, GPU, or FPGA. The specific physical component can be any component or combination of components with the above functions. The specific method depends on the actual scenario and is not limited here.

[0122] The aforementioned control device triggers a rotational motion in the robot when dragging occurs during a movement task. This movement task is supported by a first and a second line laser beam. During the rotation, a line laser beam is emitted into the target scene via a transmitting unit, and the first line laser beam reflected horizontally in the target scene is received by a receiving unit to obtain supplementary point cloud data. Based on this supplementary point cloud data, the robot's positioning is restored to perform the movement task. This achieves intelligent drag event handling. Because the robot's rotation is triggered during dragging, the line laser module can supplement point cloud data, thus enabling positioning restoration. This avoids data instability caused by the limited field of view of the line laser module and improves the stability of the robot's task execution during dragging.

[0123] Specific limitations regarding the robot's control device can be found in the limitations regarding the robot's control method described above, and will not be repeated here. Each unit module in the aforementioned robot control device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in a computer device, or stored in software in the computer device's memory, so that the processor can call and execute the corresponding operations of each module.

[0124] Another embodiment of this application also proposes a computing device, see [link to relevant documentation] Figure 11 As shown, an exemplary embodiment of this specification also provides a computing device, including: a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the steps in the robot control method according to various embodiments of this specification described above.

[0125] The internal structure of the computing device can be as follows: Figure 11 As shown, the computing device includes a processor, memory, network interface, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it follows the steps of the robot control method according to various embodiments of this specification as described in the above embodiments.

[0126] The processor may include the main processor, as well as baseband chips, modems, etc.

[0127] The memory stores a program that executes the technical solution of this invention, and may also store an operating system and other critical business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0128] The processor can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0129] Input devices may include devices that receive data and information input by the user, such as keyboards, mice, cameras, scanners, light pens, voice input devices, touch screens, pedometers, or gravity sensors.

[0130] Output devices may include devices that allow information to be output to the user, such as displays, printers, speakers, etc.

[0131] The communication interface may include any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0132] The processor executes the program stored in the memory and calls other devices, which can be used to implement the various steps of any robot control method provided in the above embodiments of this application.

[0133] The computing device may also include a display component and a voice component. The display component may be a liquid crystal display screen or an e-ink display screen. The input device of the computing device may be a touch layer covering the display component, or a button, trackball or touchpad set on the casing of the computing device, or an external keyboard, touchpad or mouse, etc.

[0134] Those skilled in the art will understand that Figure 11 The structures shown are merely block diagrams of some structures related to the solutions in this specification and do not constitute a limitation on the computing devices on which the solutions in this specification are applied. Specific computing devices may include more or fewer components than those shown in the figures, or combine certain components, or have different component arrangements.

[0135] In addition to the methods and devices described above, the robot control methods provided in the embodiments of this specification can also be computer program products, which include computer programs that, when run by a processor, cause the processor to perform the steps in the robot control methods according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0136] The computer program product described herein can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments described herein. These programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0137] Furthermore, embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor of the steps in the robot control methods according to various embodiments of this specification as described in the "Exemplary Methods" section above.

[0138] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this specification can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0139] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0140] The embodiments described above are merely illustrative of several implementation methods outlined in this specification. While the descriptions are specific and detailed, they should not be construed as limiting the scope of the solutions provided in this specification. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this specification, and these all fall within the scope of protection of this specification. Therefore, the scope of protection for this patent should be determined by the appended claims.

Claims

1. A method for controlling a robot, characterized in that, The robot is a self-moving cleaning robot, which includes a main body. A line laser module is located on the front side of the main body along the direction of travel. The line laser module includes a transmitting unit and a receiving unit. The line laser module is used to emit a first line laser beam emitted horizontally and a second line laser beam emitted at an angle downwards relative to the horizontal plane through the transmitting unit. The method includes: When the robot is dragged during the execution of a movement task, it is triggered to perform a rotation action. The movement task is supported by the first line laser beam and the second line laser beam. During the rotation operation, a line laser beam is emitted into the target scene through the transmitting unit, and the first line laser beam reflected in the target scene along the horizontal direction by the receiving unit is received to obtain supplementary point cloud data. Based on the supplementary point cloud data, the robot's positioning is restored in order to perform the movement task.

2. The method according to claim 1, characterized in that, When dragging occurs during the robot's movement task, the robot is triggered to perform a rotation action, including: During the robot's mobile task, scene point cloud data corresponding to a preset time period is acquired through the line laser module, and inertial data corresponding to the preset time period is acquired through the robot's inertial detection module. The scene point cloud data is compared with the memory map corresponding to the target scene to obtain the first abnormal parameter, and the second abnormal parameter is obtained by determining the fluctuation amplitude corresponding to the inertial data. When the first abnormal parameter and the second abnormal parameter indicate that the robot is being dragged, the robot is triggered to perform a rotation action.

3. The method according to claim 2, characterized in that, During the robot's movement task, the process involves acquiring scene point cloud data corresponding to a preset time period via the line laser module and acquiring inertial data corresponding to the preset time period via the robot's inertial detection module, including: During the process of the robot performing a movement task in the target scene, the initial time parameters corresponding to the robot are obtained; Determine the scene time parameters corresponding to the target scene; The initial time parameter is configured based on the scenario time parameter, so as to determine the preset time period according to the configured initial time parameter; The robot acquires scene point cloud data corresponding to the preset time period through its line laser module, and acquires inertial data corresponding to the preset time period through its inertial detection module.

4. The method according to claim 3, characterized in that, The step of configuring the initial time parameter based on the scene time parameter, and determining the preset time period according to the configured initial time parameter, includes: Obtain the size parameters corresponding to the target scene, and the reference size corresponding to the scene time parameters; The size of the area where the robot is located in the scene is determined based on the size parameters; The adjustment parameters are determined based on the ratio of the region size to the reference size; The scene time parameter is adjusted by the adjustment parameter, and the initial time parameter is configured based on the adjusted scene time parameter, so as to determine the preset time period according to the configured initial time parameter.

5. The method according to claim 2, characterized in that, The step of comparing the scene point cloud data with the memory map corresponding to the target scene to obtain a first anomaly parameter, and obtaining a second anomaly parameter by determining the fluctuation amplitude corresponding to the inertial data, includes: The maps at different acquisition times in the scene point cloud data are compared with the memory map corresponding to the target scene to obtain the point cloud matching score corresponding to different acquisition times. The point cloud matching score is compared with the matching threshold to identify abnormal point cloud data. The first abnormal parameter is determined based on the proportion of the abnormal point cloud data to the scene point cloud data; The detected data points in the inertial data are compared with preset inertial parameters to determine the fluctuation amplitude; The fluctuation amplitude is compared with the fluctuation threshold to identify inertial anomaly data; The second anomaly parameter is determined based on the proportion of the inertial anomaly data to the total inertial data.

6. The method according to claim 5, characterized in that, The step of comparing the point cloud matching score with a matching threshold to determine abnormal point cloud data includes: Obtain the matching threshold corresponding to the target scene; Based on the distribution of obstacle elements in the memory map, the scene complexity corresponding to the area where the robot is located is determined; The matching threshold is adjusted according to the complexity of the scenario; The point cloud matching score is compared with the adjusted matching threshold to identify abnormal point cloud data.

7. The method according to claim 5, characterized in that, The step of comparing the fluctuation amplitude with the fluctuation threshold to determine inertial anomaly data includes: Obtain the robot's historical inertial data in the target scene; Determine the fluctuation reference range indicated by the historical inertial data; Configure the fluctuation threshold according to the fluctuation reference range; The fluctuation amplitude is compared with the fluctuation threshold to determine the inertial anomaly data.

8. The method according to claim 2, characterized in that, When the first abnormal parameter and the second abnormal parameter indicate that the robot is being dragged, triggering the robot to perform a rotation action includes: If the first abnormal parameter collected by the linear laser module and the second abnormal parameter collected by the inertial detection module both meet the preset conditions, then the dragging status signal is issued. Based on the drag status signal, the robot is controlled to continuously update the first abnormal parameter and the second abnormal parameter, and the scene map is prohibited from being updated during the drag status signal output process; If neither the updated first abnormal parameter nor the updated second abnormal parameter meets the preset conditions, the robot is triggered to perform the rotation action in place.

9. The method according to claim 8, characterized in that, If both the first abnormal parameter acquired by the linear laser module and the second abnormal parameter acquired by the inertial detection module meet preset conditions, then the dragging state signal is emitted, including: If the first abnormal parameter collected by the linear laser module and the second abnormal parameter collected by the inertial detection module both meet the preset conditions, a suspected dragging signal will be issued. The robot is triggered to perform a rotation action based on the suspected dragging signal; During the rotational motion, the third abnormal parameter acquired by the line laser module is controlled, and the fourth abnormal parameter acquired by the inertial detection module is also controlled. If both the third abnormal parameter and the fourth abnormal parameter meet the preset conditions, then the drag status signal is issued.

10. A cleaning device, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the robot control method according to any one of claims 1 to 9.

Images