Intelligent warehouse equipment state information acquisition method and system

Abstract

The application provides a kind of intelligent warehousing equipment state information acquisition method and system, related to intelligent warehousing technical field, and the intelligent warehousing equipment state information acquisition method includes: obtaining the initial state data of the intelligent warehousing equipment collected by multi-modal sensor group, wherein the multi-modal sensor group at least includes laser scanner, ultra-wide band positioning base station and equipment vibration sensor;The initial state data is dynamically estimated, and the predicted trajectory of the intelligent warehousing equipment is obtained;According to the predicted trajectory, the scanning path of the laser scanner is generated;Intelligent warehousing equipment is scanned based on the scanning path, and the intelligent warehousing equipment state information is obtained.The intelligent warehousing equipment state information acquisition method and system of the application solve the technical problems of high equipment state update delay and low acquisition efficiency caused by polling collection and PLC dependence in related technologies.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent warehousing, and in particular to an intelligent warehousing equipment state information acquisition method and system. BACKGROUND

[0002] With the continuous improvement of the intelligent level of the automatic warehousing system, its advantages in reducing labor costs, improving operation safety and management efficiency are increasingly prominent, and it has been widely used in manufacturing, e-commerce logistics and other fields. In the running process of the intelligent warehousing system, it is often integrated and interacted with various intelligent devices such as stacker, automatic guided vehicle (AGV), conveying line, etc. Among them, the host computer software needs to acquire the device running state data in real time to realize the functions of operation visualization, abnormal alarm and warehousing management optimization, etc. However, the existing device state acquisition method still has some technical bottlenecks, which restricts the further improvement of the overall efficiency and response real-time of the system.

[0003] Currently, the common state acquisition methods mainly have the following defects: First, the polling-based data acquisition mechanism is inefficient, resulting in high device state update delay. This method lacks stability in acquisition timing and is highly dependent on the programmable logic controller (PLC) write operation to the data block (DB). Since the time required for each acquisition is significantly affected by the size of the DB data, an increase in data size will directly lead to an increase in acquisition cycle, thereby reducing the acquisition efficiency. In addition, changes in device state need to be written into the DB by the PLC through sensors or motor parameters after sensing abnormalities, and then read by the upper computer according to the polling cycle. This process introduces multi-level waiting and transmission delay, further exacerbating the lag of state information. Second, the processing load of video monitoring image data is huge, making it difficult to achieve real-time analysis and response. Taking a typical 1080P camera as an example, the data volume of a single frame of image is about 6.22 MB. If calculated at a frame rate of 30 frames per second, a single video stream generates about 186.6 MB of data per second, and accumulates about 15.4 TB per day. Image data processing usually includes multiple stages such as raw image acquisition, preprocessing, target detection, feature extraction, and state analysis, each of which involves a large amount of calculation: for example, when performing median filtering and other denoising operations in the preprocessing stage, a single frame of image needs to perform about 124 million multiplication operations; when using YOLOv5 and other models for target detection, the calculation amount of each frame of image is about 10^10 floating point operations (FLOPs), even if using high-performance GPUs such as RTX 3080, the processing speed of a single video stream can only reach about 80 frames per second; and in the feature extraction stage, using ResNet-50 and other networks to process each identified target requires about 4.1×10^9 FLOPs. In addition, the dense arrangement of devices such as shelves, stackers, and AGVs in the warehouse environment causes occlusion problems, which often require the introduction of 3D pose estimation to achieve accurate identification, increasing the calculation amount by about 10 times; dynamic lighting conditions also require the use of HDR fusion technology, resulting in a data volume of 3 times that of the original; at the same time, the detection of small-sized device parts (such as targets smaller than 50 pixels) often relies on super-resolution reconstruction, resulting in an increase in computing power consumption by about 8 times. These factors together exacerbate the real-time technical challenges of the video monitoring system. Third, the deployment structure of the sensor network is complex, and the overall implementation and maintenance cost is high. To effectively cover a warehouse area of 100m x 50m, it is usually necessary to deploy at least 50 sensor nodes. Due to the shielding effect of metal shelves on wireless signals, signal attenuation can reach more than 20 dB, and about 15% of additional relay nodes are often needed to ensure communication quality.In terms of related hardware and network facilities, including industrial switches (about $500 each, at least 5 units), fiber backbone network (about $100 / m), etc.; edge computing nodes usually need to be configured with industrial computers and GPU devices, and each set costs about $5000; in addition, the software license fee of the supervisory control and data acquisition system (SCADA) is about $150 per point, and at a scale of 50 points, the license fee alone is $7500. Comprehensive estimates show that the total cost of deploying a system covering 5000 square meters of warehouse area is between $120,000 and $250,000, and does not yet include subsequent operation and maintenance expenses.

[0004] Therefore, the prior art needs to be further developed. SUMMARY

[0005] The purpose of the present application is to overcome the above technical deficiencies and provide an intelligent warehouse equipment state information acquisition method and system to solve the technical problems of high equipment state update delay and low acquisition efficiency caused by polling acquisition and PLC dependence in related technologies.

[0006] To achieve the above technical purpose, the present application adopts the following technical scheme: an intelligent warehouse equipment state information acquisition method is provided, comprising: acquiring initial state data of intelligent warehouse equipment collected by a multi-modal sensor group, wherein the multi-modal sensor group at least includes a laser scanner, an ultra-wideband positioning base station and a device vibration sensor; dynamically estimating the initial state data to obtain a predicted trajectory of the intelligent warehouse equipment; generating a scanning path of the laser scanner according to the predicted trajectory; scanning the intelligent warehouse equipment based on the scanning path to obtain the intelligent warehouse equipment state information.

[0007] Further, the intelligent warehouse equipment state information acquisition method further comprises: performing time synchronization processing on the initial state data from the multi-modal sensor group; converting the time-synchronized data to a preset global coordinate system to obtain observation data.

[0008] Further, the intelligent warehouse equipment state information acquisition method further comprises: establishing an observation function based on the observation data; calculating the Jacobian matrix of the observation function at the predicted state vector of the current time to obtain an observation matrix.

[0009] Further, the method for obtaining the predicted trajectory of the intelligent warehouse equipment comprises: constructing a state vector based on the initial state data, wherein the state vector at least includes the position, speed and orientation angle of the intelligent warehouse equipment; dynamically estimating the state vector based on the Kalman filter algorithm to generate an optimal estimate of the state vector of the intelligent warehouse.

[0010] Further, the method for dynamically estimating the state vector based on the Kalman filtering algorithm comprises: obtaining a predicted state vector and a corresponding predicted covariance based on a state transition model; and correcting the predicted state vector at the current time by using the Kalman gain to obtain an optimal estimation of the state vector.

[0011] Further, the state vector is specifically (p x ,p y ,v x ,v y ,θ) T , wherein p x ,p y are position coordinates of the intelligent warehousing equipment in a two-dimensional plane, v x ,v y are movement speeds of the intelligent warehousing equipment in x and y directions, and θ is an orientation angle of the intelligent warehousing equipment.

[0012] Further, the method for generating the scanning path of the laser scanner comprises: determining a scanning area that needs to be covered by the laser scanner based on the predicted trajectory; and dynamically adjusting a scanning resolution according to a real-time running speed of the intelligent warehousing equipment on the preset trajectory, wherein the real-time running speed is negatively correlated with the scanning resolution.

[0013] Further, the intelligent warehousing equipment state information acquisition method further comprises: jointly diagnosing a device mechanical state of the intelligent warehousing equipment based on vibration data measured by the device vibration sensor and profile data collected by the laser scanner; and recording accurate position information provided by the ultra-wideband positioning base station when the vibration data exceeds a preset threshold and the profile data indicates that the device has deformed.

[0014] Further, the intelligent warehousing equipment state information acquisition system comprises: a data acquisition unit configured to acquire initial state data of an intelligent warehousing equipment collected by a multi-modal sensor group, wherein the multi-modal sensor group at least comprises a laser scanner, an ultra-wideband positioning base station and a device vibration sensor; a dynamic prediction unit configured to dynamically estimate the initial state data to obtain a predicted trajectory of the intelligent warehousing equipment; a path planning unit configured to generate a scanning path of the laser scanner according to the predicted trajectory; and a device scanning unit configured to scan the intelligent warehousing equipment based on the scanning path to obtain the intelligent warehousing equipment state information.

[0015] The application further provides a computer readable storage medium, wherein computer readable instructions are stored on the computer readable storage medium, and the computer readable instructions are used to realize each step of the intelligent warehouse equipment state information collection method or the intelligent warehouse equipment state information collection system.

[0016] Beneficial effects:

[0017] 1. The intelligent warehouse equipment state information collection method constructs a multi-modal sensor group including a laser scanner, an ultra-wideband positioning base station and a device vibration sensor, and collects spatial profile data of the laser scanner, centimeter-level position data of the ultra-wideband positioning base station and mechanical state data of the device vibration sensor, and then, after a series of data preprocessing, the three types of heterogeneous raw observation data are combined into an observation vector of Kalman filtering under a unified time stamp and coordinate system, and based on the iteration process of the prediction-update in the Kalman filtering algorithm, the multi-source data are optimally fused and dynamically estimated to generate a predicted trajectory including position, speed and motion trend, and then the laser scanner is driven to perform adaptive scanning according to the predicted trajectory, which completely changes the traditional polling collection mechanism, effectively solves the technical problems of high equipment state update delay and low collection efficiency caused by polling collection and PLC dependence in the prior art, and realizes real-time, efficient and accurate perception of the running state of the intelligent warehouse equipment.

[0018] 2. The intelligent warehouse equipment state information collection method can accurately diagnose and cross-verify the mechanical state of the equipment by deeply fusing and jointly analyzing the device vibration sensor data, the laser scanner profile data and the ultra-wideband precise positioning data in the Kalman filtering framework, effectively distinguishes transient interference from real faults, thereby greatly reduces the system false alarm rate by about 84.7%, and realizes millimeter-level accurate positioning of faults, thereby greatly enhancing the safety and stability of the warehouse system operation.

[0019] 3. The intelligent warehouse equipment state information collection method generates and adjusts the laser scanning path in real time according to the predicted trajectory, and the system can focus on scanning the key areas, and appropriately reduces the scanning density for non-key areas or high-speed moving equipment. This adaptive scanning mechanism reduces the video monitoring image data processing amount by about 40% under the premise of ensuring the state monitoring effect, and reduces the unnecessary consumption of the computing power of the edge computing node. This not only reduces the limit requirement for hardware computing power, saves the system deployment and operation and maintenance cost, but also solves the problem of large video data processing amount and difficulty in real-time response in the background technology. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flow chart of the intelligent warehouse equipment state information collection method adopted by the embodiment of the application.

[0021] Figure 2 is the structural schematic diagram of the intelligent warehouse equipment state information acquisition system adopted by the embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable personnel in the art to better understand the present application scheme, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor should belong to the scope of protection of the present application.

[0023] According to the embodiment of the present application, an intelligent warehouse equipment state information acquisition method is provided, please refer to Figures 1-2 , comprising:

[0024] S100 acquiring initial state data of intelligent warehouse equipment collected by a multi-modal sensor group, wherein the multi-modal sensor group at least includes a laser scanner, an ultra-wideband positioning base station and a device vibration sensor;

[0025] Specifically, the laser scanner sets the scanning parameters according to the distance formula d=c·Δt / 2, wherein c=3×10 8 m / s, Δt=transmission to reception time difference, scanning frequency: 75 Hz (75 times full area scanning per second), angular resolution: 0.1°, scanning angle: 190°.

[0026] Preferably, the laser scanner adopts LMS511 series of SICK China, scanning angle 190°, scanning frequency 50 Hz, accuracy ±2mm. Each scanner covers a fan-shaped area with a radius of 30 meters.

[0027] It should be noted that the UWB (Ultra-Wideband, Ultra-Wideband) positioning base station transmits and receives nanosecond-level extremely narrow pulse signals, and realizes high-precision positioning (up to centimeter level) by using signal flight time (ToF) and time difference of arrival (TDoA) algorithms.

[0028] Specifically, the bandwidth of the ultra-wideband pulse is ≥500MHz, the pulse width is extremely short (nanosecond level), and it has strong anti-interference ability. At the same time, because the signal is extremely narrow, the time measurement accuracy can reach picosecond level (1 picosecond=10⁻¹² second), so it has high time resolution.

[0029] The UWB working steps (taking TDoA as an example) include:

[0030] (1) Tag transmits signals: UWB tags periodically broadcast pulse signals.

[0031] (2) Base station receives signals: At least 3 base stations record signal arrival times (time synchronization is required).

[0032] (3) Calculate time difference: The central processor compares the time difference received by each base station (e.g. base station A is 50ns earlier than base station B).

[0033] (4) Solve position: Determine the tag coordinates by the intersection algorithm of hyperbolas.

[0034] In this embodiment, the equipment vibration sensor utilizes the piezoelectric effect to cause the piezoelectric crystal to generate a change in electric charge, and outputs a voltage signal through a signal amplifier. Its working principle is: Q = d * F, where Q = generated charge, d = piezoelectric constant, and F = applied force.

[0035] Preferably, the vibration sensor uses Bonner China QM42VT, with a range of ±50g and a frequency response range of 5-10kHz, and is transmitted through the IO-Link protocol (a point-to-point communication interface protocol for intelligent sensors / actuators). The technical parameters of the equipment vibration sensor of this embodiment are shown in Table 1:

[0036] Table 1: Technical parameters of the equipment vibration sensor

[0037] Parameter Indicator Industrial significance Range ±50g Detectable severe impact Frequency response range 5Hz-10kHz Cover bearing / gear fault characteristic frequency Resolution 0.0005g Early micro-vibration detection Operating temperature -40℃~125℃ Adapt to cold storage / high temperature environment

[0038] This embodiment reduces the false alarm rate by 84.7% through multi-modal sensor fusion, using vibration + laser deformation + UWB pulse precise positioning, and joint diagnosis. The fault positioning accuracy is millimeter level.

[0039] In specific practice, the intelligent storage equipment state information acquisition method further includes preprocessing of initial state data.

[0040] Specifically, the initial state data from the multi-modal sensor group is subjected to time synchronization processing; and the data subjected to time synchronization processing is converted to a preset global coordinate system to obtain observation data.

[0041] Further, an observation function is established based on the observation data; and a Jacobian matrix of the observation function at a predicted state vector of the current time instant is calculated to obtain an observation matrix.

[0042] Wherein, the observation function is defined as h(x) = y - Hx ,

[0043] Calculate the Jacobian matrix H j , H j is the partial derivative matrix of h(x) with respect to state x:

[0044] H j= ,

[0045] Specifically, , , , and further obtain a linearized H matrix: H= wherein, and are unit direction vectors (from the AGV to the landmark).

[0046] In this way, by performing time synchronization processing on the initial state data from the multi-modal sensor group, the data timing misalignment problem caused by the differences in collection frequency and transmission delay of different sensors is effectively solved, a unified time reference is established for subsequent data fusion, the effectiveness of multi-source data fusion is ensured from the source, the data converted to the preset global coordinate system after time synchronization processing eliminates the space description inconsistency problem caused by the coexistence of multiple coordinate systems such as the local coordinate system of the laser scanner and the global coordinate system of UWB, and provides a unified space reference framework for all observation data. By establishing an observation function and calculating the Jacobian matrix thereof at the current predicted state vector, an observation matrix is obtained, which successfully converts the actual nonlinear observation relationship such as laser ranging into a linear form that can be directly processed by the Kalman filtering algorithm, overcomes the calculation complexity of directly using a nonlinear model, and maximizes the accuracy of the model, thereby providing core algorithm support for generating a high-precision predicted trajectory.

[0047] S200 performs dynamic estimation on the initial state data to obtain a predicted trajectory of the intelligent warehouse device;

[0048] Specifically, the method for obtaining the predicted trajectory of the intelligent warehouse device includes:

[0049] constructing a state vector based on the initial state data, wherein the state vector at least includes the position, speed and orientation angle of the intelligent warehouse device;

[0050] In some embodiments, the state vector is specifically [p x ,p y ,v x ,v y ,θ] T wherein p x p y are the position coordinates of the intelligent warehouse device in a two-dimensional plane, v x v y are the motion speeds of the intelligent warehouse device in the x and y directions, and θ is the orientation angle of the intelligent warehouse device.

[0051] The state vector is dynamically estimated based on a Kalman filtering algorithm to generate an optimal estimation of the state vector of the intelligent warehouse.

[0052] The method of dynamically estimating the state vector based on the Kalman filtering algorithm comprises:

[0053] Based on the state transition model, a predicted state vector and its corresponding predicted covariance are obtained.

[0054] The predicted state vector at the current time is corrected using the Kalman gain to obtain the optimal estimation of the state vector.

[0055] Embodiment one:

[0056] This embodiment takes the positioning and navigation of warehouse AGVs as an example, taking the position, speed, and direction angle of the AGV as state variables, using a motion model (such as encoder or IMU data) to predict the state of the AGV at the next time, and using an observation model (such as the position matched by the laser radar, the position of the visual feature point, and the UWB ranging information) to correct the predicted state.

[0057] Specifically, in this embodiment, the state vector is used to describe the real state hidden inside the system, such as the coordinates (p x ,p y ) and the heading angle θ of the AGV.

[0058] The observation vector z is the original data actually measured by the sensor, such as the distance d measured by the laser radar to the landmark.

[0059] Since the Kalman filter requires a linear model, but the actual physical relationship is mostly nonlinear, an observation equation (or measurement equation) needs to be introduced, specifically z=H·x+noise, where H linearly maps the state space to the observation space, x is the current state, and z is the value measured by the sensor under the current state. The observation matrix H can establish a mathematical relationship between the state x and the observation value z.

[0060] The calculation of the landmark distance in AGV positioning is specifically:

[0061] AGV state: x= (position and orientation);

[0062] Landmark position: known fixed coordinates (l x ,l y );

[0063] Sensor: the Euclidean distance d measured by the laser radar to the landmark.

[0064] The method of calculating the relative position is specifically: the relative coordinates of the landmark in the AGV coordinate system are: ,

[0065] The method of converting relative coordinates into polar coordinates is:

[0066] The theoretical distance d and the relative angle f are:

[0067] ,

[0068] f = arctan2 (Ay, Ax) - q,

[0069] Considering the AGV orientation, if the laser radar directly outputs the distance d, the observation equation is:

[0070]

[0071] At this time, the observation vector is a scalar: z = [d].

[0072] In an embodiment, the generation of the observation matrix H specifically includes:

[0073] Define a nonlinear function h(x) = ,

[0074] Calculate the Jacobian matrix H j , H j is the partial derivative matrix of h(x) with respect to the state x:

[0075] H j= ,

[0076] Specifically, , , , and then get the linearized H matrix: H = , wherein and are unit directional vectors (from the AGV to the landmark).

[0077] It should be noted that the linearized H matrix is essentially the sensitivity (directional derivative) of the distance change to the position change.

[0078] Substitute z = H x + noise, where x is the current state, and z is the value measured by the sensor under the current state. In this way, the current state can be calculated according to the data measured by the sensor.

[0079] S300 generates a scanning path of the laser scanner according to the predicted trajectory;

[0080] The method of generating the scanning path of the laser scanner includes:

[0081] determine a scanning area that needs to be covered by the laser scanner based on the predicted trajectory;

[0082] adjust the scanning resolution dynamically according to a real-time running speed of the intelligent storage device on the preset trajectory, wherein the real-time speed is negatively correlated with the scanning resolution.

[0083] In this way, the scanning line path is dynamically adjusted according to the device motion trajectory prediction model, and the fault-prone area is preferentially scanned, so that efficient optimization of system calculation and scanning resources is achieved. The traditional fixed-frequency scanning mode will cause waste of computing power when the device is stationary or running at low speed, and may cause state omission due to insufficient sampling when the device is running at high speed. The embodiment establishes a negative correlation between the real-time running speed of the device and the scanning resolution, so as to realize dynamic on-demand allocation of scanning resources. When the device is running at high speed, the resolution is automatically reduced to ensure full coverage, and when the device is running at low speed or is stationary, the resolution is increased for fine detection. This adaptive mechanism effectively reduces the laser point cloud data processing amount by about 35% under the premise of ensuring monitoring effect, and significantly reduces the system computing load.

[0084] Meanwhile, the scanning area determined based on the predicted trajectory enables the system to prospectively cover the motion path of the device at a future time, and compared with the traditional fixed-area scanning mode, the scanning mode of the embodiment avoids invalid scanning of areas without the device, while ensuring key monitoring of areas to be passed through by the device, so that the system can obtain accurate data of the key areas before a fault occurs, thereby providing a data basis for early fault diagnosis.

[0085] Further, the embodiment realizes the transformation from "passive response" to "active adaptation" by closed-loop integration of trajectory prediction, speed sensing and scanning control, so that the system can intelligently adjust the monitoring strategy according to the actual motion state of the device, ensuring continuous tracking of the trajectory under high-speed motion, and realizing fine detection of key parts, thereby significantly improving the overall monitoring efficiency and reliability while reducing the system resource consumption.

[0086] In specific practice, the comparison results of the intelligent storage device state information acquisition system of the present application and the traditional polling scheme are shown in Table 2.

[0087] Table 2: Comparison of intelligent storage device state information acquisition system and traditional polling scheme

[0088] Indicator Traditional polling scan The present invention Lift range Fault discovery delay 320±50ms 85±15ms 73.4% Scan energy consumption 100% 63% 37% High priority area coverage 76% 98% 22%

[0089] The comparison results of the intelligent storage device state information acquisition system of the present application and the video monitoring scheme are shown in Table 3.

[0090] Table 3: Comparison of intelligent storage device state information acquisition system and video monitoring scheme

[0091] Indicator Video monitoring scheme The present invention Advantage multiple Single device data volume 186.6MB / s (1080P@30fps) 0.8 MB / s (feature vector) 233× Processing delay 85-360 ms 8-15 ms 10× Computing power requirement 45 TOPS (4K) 0.5 TOPS 90× Network load 1.2 Gbps / road 6.4 Mbps / road 187×

[0092] The comparison results of the intelligent warehouse equipment state information acquisition system and the sensor scheme of the present application are shown in Table 4.

[0093] Table 4: Comparison of intelligent warehouse equipment state information acquisition system and sensor scheme

[0094] Cost items Sensor scheme The present invention Saving Hardware procurement $180,000 $65,000 $115,000 Installation engineering $50,000 $12,000 $38,000 Maintenance and calibration $75,000 $18,000 $57,000 Energy consumption $14,000 $3,800 $10,200 Total $319,000 $98,800 $220,200

[0095] According to Tables 2-4, using the intelligent warehouse equipment state information acquisition scheme of the present application, the acquisition delay is reduced from ≥500ms to <100ms, the data processing amount is reduced by 70% (compared with traditional image processing), and the equipment fault recognition accuracy is improved to 99.2%, greatly improving the acquisition efficiency, reducing the data calculation amount, effectively solving the technical problems of high equipment state update delay and low acquisition efficiency caused by polling acquisition and PLC dependence in the prior art, and realizing real-time, efficient and accurate perception of the running state of the intelligent warehouse equipment.

[0096] S400 scans the intelligent warehouse equipment based on the scanning path to obtain the intelligent warehouse equipment state information.

[0097] It should be noted that after obtaining the intelligent warehouse equipment state information, the fault diagnosis component will judge the state information.

[0098] Specifically, based on the vibration data measured by the equipment vibration sensor and the profile data collected by the laser scanner, the mechanical state of the intelligent warehouse equipment is jointly diagnosed.

[0099] When the vibration data exceeds the preset threshold and the profile data indicates that the equipment has deformed, the accurate position information provided by the ultra-wideband positioning base station is recorded.

[0100] The embodiment also provides an intelligent warehouse equipment state information acquisition system, which comprises: a data acquisition unit, the data acquisition unit is used to acquire initial state data of intelligent warehouse equipment collected by a multi-modal sensor group, wherein the multi-modal sensor group at least includes a laser scanner, an ultra-wideband positioning base station and an equipment vibration sensor; a dynamic prediction unit, the dynamic prediction unit is used to dynamically estimate the initial state data to obtain a predicted trajectory of the intelligent warehouse equipment; a path planning unit, the path planning unit is used to generate a scanning path of the laser scanner according to the predicted trajectory; and a device scanning unit, the device scanning unit is used to scan the intelligent warehouse equipment based on the scanning path to obtain the intelligent warehouse equipment state information.

[0101] With the above settings, the data acquisition unit forms an observation vector of Kalman filtering by constructing a multi-modal sensor group containing the laser scanner, the ultra-wideband positioning base station and the equipment vibration sensor, and directly forming three types of heterogeneous raw observation data, i.e., the spatial profile data collected by the laser scanner, the centimeter-level position data acquired by the ultra-wideband positioning base station and the mechanical state data measured by the equipment vibration sensor, under a unified time stamp and coordinate system, the dynamic prediction unit optimally fuses and dynamically estimates these multi-source data based on the iteration process of "prediction-update" in the Kalman filtering algorithm, generates a predicted trajectory containing position, velocity and motion trend, and then the path planning unit and the equipment scanning unit drive the laser scanner to perform adaptive scanning according to the predicted trajectory, which completely changes the traditional polling acquisition mechanism, effectively solves the technical problems of high equipment state update delay and low acquisition efficiency caused by polling acquisition and PLC dependence in the prior art, and realizes real-time, efficient and accurate perception of the running state of the intelligent warehousing equipment.

[0102] The embodiment also provides a computer readable storage medium, which stores computer readable instructions, and the computer readable instructions realize each step of the intelligent warehousing equipment state information acquisition method or the intelligent warehousing equipment state information acquisition system when executed by a processor.

[0103] The embodiment of the present application can be in the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing program codes. The computer readable storage medium includes permanent and non-permanent, removable and non-removable media, and can be realized by any method or technology. Information storage. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer readable storage media include but are not limited to: new memory such as phase change memory / resistance change memory / magnetic memory / ferroelectric memory (PRAM / RRAM / MRAM / FeRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information that can be accessed by a computing device.

[0104] It should be noted that the terms "first", "second" and the like in the description and in the claims of the present application are used to distinguish between similar objects and not necessarily describe a particular sequential or chronological order. It is to be understood that the use of these terms in the description and in the claims of the present application is to be construed to cover not only the particular object so named but also its functional equivalents unless the context clearly indicates otherwise. Furthermore, the terms "comprising", "having", "including", and the like, when used in the present specification and in the following claims, are each intended to denote an open-ended term such that the listed steps or components do not exclude additional steps or components. The terms "comprise", "comprising", "comprises", "including", "include", "includes" and / or "contain" or variations thereof are used herein to indicate the presence of stated features, steps, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0105] Optionally, the specific examples in the embodiments can refer to the examples described in the above embodiments, and the embodiments will not be described here again.

[0106] The serial numbers of the above embodiments of the present application are only for description, not representing the advantages and disadvantages of the embodiments.

[0107] In the above embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0108] The above only describes the preferred embodiments of the present application, and it should be noted that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should be considered as the protection scope of the present application.

Claims

1. A method for collecting intelligent warehouse equipment state information, characterized in that, The method comprises the following steps: The initial state data of the intelligent warehousing equipment collected by a multi-modal sensor group is acquired, wherein the multi-modal sensor group at least includes a laser scanner, an ultra-wideband positioning base station and a device vibration sensor, and the intelligent warehousing equipment state information acquisition method further comprises: performing time synchronization processing on the initial state data from the multi-modal sensor group; converting the data subjected to the time synchronization processing to a preset global coordinate system to obtain observation data, establishing an observation function based on the observation data, and the observation function is defined as h(x)= wherein (lx,ly) is a defined anchor point coordinate in a map; (px,py) is a coordinate of the AGV in the map; a Jacobian matrix of the observation function at a predicted state vector of a current moment is calculated to obtain an observation matrix; dynamically estimating the initial state data to obtain a predicted trajectory of the intelligent warehousing device, wherein the method for obtaining the predicted trajectory of the intelligent warehousing device comprises the following steps: constructing a state vector based on the initial state data, wherein the state vector at least comprises the position, speed and orientation angle of the intelligent warehousing device; dynamically estimating the state vector based on a Kalman filtering algorithm to generate an optimal estimation of the state vector of the intelligent warehousing device; generating a scanning path of the laser scanner according to the predicted trajectory; scanning the intelligent warehousing device based on the scanning path to obtain the intelligent warehousing device state information; jointly diagnosing the device mechanical state of the intelligent warehousing device based on the vibration data measured by the device vibration sensor and the profile data collected by the laser scanner; when the vibration data exceeds a preset threshold and the profile data indicates that the device has deformed, recording the accurate position information provided by the ultra-wideband positioning base station. 2.The intelligent warehouse equipment state information collection method according to claim 1, characterized in that, The method for dynamically estimating the state vector based on the Kalman filtering algorithm comprises the following steps: obtaining a predicted state vector and its corresponding predicted covariance based on a state transition model; correcting the predicted state vector at the current time by using a Kalman gain to obtain the optimal estimation of the state vector. 3.The intelligent warehouse equipment state information collection method of claim 1, characterized in that, The state vector is specifically (p x ,p y ,v x ,v y ,θ) T , wherein p x ,p y are position coordinates of the intelligent warehousing equipment in a two-dimensional plane, v x ,v y are movement speeds of the intelligent warehousing equipment in x and y directions, and θ is an orientation angle of the intelligent warehousing equipment. 4.The intelligent warehouse equipment state information collection method of claim 1, characterized in that, The method for generating the scanning path of the laser scanner comprises the following steps: determining the scanning area that needs to be covered by the laser scanner based on the predicted trajectory; dynamically adjusting the scanning resolution according to the real-time running speed of the intelligent warehousing device on the predicted trajectory, wherein the real-time running speed is negatively correlated with the scanning resolution. 5.A system for collecting intelligent warehouse equipment state information, configured to implement the method for collecting intelligent warehouse equipment state information according to any one of claims 1-4, characterized in that, The intelligent warehousing device state information acquisition system comprises: a data acquisition unit, which is used to acquire the initial state data of the intelligent warehousing device collected by a multi-modal sensor group, wherein the multi-modal sensor group at least comprises a laser scanner, an ultra-wideband positioning base station and a device vibration sensor; a dynamic prediction unit, which is used to dynamically estimate the initial state data to obtain a predicted trajectory of the intelligent warehousing device; a path planning unit, which is used to generate a scanning path of the laser scanner according to the predicted trajectory; a device scanning unit, which is used to scan the intelligent warehousing device based on the scanning path to obtain the intelligent warehousing device state information.

6. A computer-readable storage medium having stored thereon computer-readable instructions, wherein, The computer readable instructions are executed by the processor to realize the steps of the intelligent warehousing device state information acquisition method in any one of claims 1-4.

Images