An AMR navigation compensation method and system for an electric mobile shelf parking deviation

Abstract

The present application relates to the technical field of warehouse equipment control and mobile robot cooperative navigation, and discloses an electric mobile rack parking deviation AMR navigation compensation method and system, which comprises the following steps: acquiring a theoretical stop position, driving the electric mobile rack to displace to form a target operation channel, collecting data in stages and dynamically adjusting the fusion weight of the multi-source detection device according to the operation stage, calculating the actual stop position and the channel boundary, calculating the deviation, converting the actual channel geometric information into compensation navigation parameters, and writing the compensation navigation parameters into the obstacle layer, the inflation layer and the no-entry layer of the AMR local cost map, controlling the AMR to scan the entrance contour and permitting entry when the comparison is consistent, and dynamically eliminating the influence of mechanical inertia and displacement deviation on geometric space, thereby improving the passing safety and docking accuracy of narrow channels.

Description

Technical Field

[0001] This invention relates to an AMR navigation compensation method and system for parking deviation of electric mobile shelving, belonging to the field of warehouse equipment control and mobile robot collaborative navigation technology. Background Technology

[0002] In dense storage systems, a collaborative operation mode of mobile shelves and automated mobile robots is commonly used. The mobile shelves move along ground rails or flat ground to create temporary working channels, and the automated mobile robots drive into the channels according to a preset digital map to complete the picking and placing of goods. In actual engineering operation, the mobile shelves are affected by the inertia of the drive motor, the clearance of the track fit, the difference in load distribution, and the fit tolerance of the mechanical locking mechanism. The actual physical position and posture of the mobile shelves when they stop inevitably deviate from the theoretical design coordinates. Existing control systems usually only confirm the stop of the shelves through limit switches or position status signals and send fixed channel center lines and entrance coordinates to the automated mobile robots. They fail to quantify the physical space changes caused by the stopping deviation. When the shelf shifts position or tilts at an angle, the actual effective width of the channel is reduced, resulting in the robot's entrance positioning being biased to one side or insufficient safety distance between the vehicle body and the shelf boundary, which can easily lead to driving stops, scraping collisions, or mechanical docking failures.

[0003] To mitigate the physical safety risks caused by positioning deviations, conventional engineering improvement approaches mainly focus on enhancing the rigidity of the shelving's mechanical positioning or relying on robot-mounted sensors for obstacle avoidance. Improving mechanical positioning accuracy requires significantly increasing the deployment cost of high-precision limiting and locking mechanisms, and cannot offset the mechanical wear and structural deformation caused by long-term operation. Relying solely on robot-mounted sensors for autonomous obstacle avoidance can easily lead to misjudging narrowed shelving sidewalls as obstacles, causing navigation protection malfunctions and frequent shutdowns. Existing methods have the following main shortcomings: 1. They only transmit static preset coordinates to the mobile robot, failing to quantify the changes in the spatial geometry of the passageway caused by mechanical motion errors; 2. Relying on unilateral autonomous obstacle avoidance by the robot easily leads to navigation protection malfunctions and frequent shutdowns in narrow dynamic passageways; 3. Relying on rigid hardware limits cannot resist cumulative wear and load displacement caused by long-term operation, creating a fundamental contradiction between the uncertainty of the physical posture of the moving structural components and the mobile robot's requirement for precise spatial boundary constraints.

[0004] Therefore, how to accurately obtain the actual physical deviation of the mobile shelf's stopping position and reconstruct it in real time into a dynamic spatial navigation constraint that the mobile robot can adapt to becomes the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: A method for compensating for parking deviation of an electric mobile shelving unit using AMR navigation, comprising the following steps: Step S1: Obtain the theoretical stopping positions of the target shelf and the target storage location; Step S2: Drive the electric mobile shelf to move and form the target operation channel; Step S3: Collect rack position data and aisle boundary data in stages using a multi-source detection device. Dynamically adjust the fusion weight of the multi-source detection device according to the operation stage of the electric mobile rack. Increase the weight of encoder and drive current sensor in the start-up and movement stages, increase the weight of locking pin status sensor in the near-stop stage, and increase the weight of lidar and visual marker in the post-stop verification stage. Calculate the actual stop position, actual aisle width, and aisle boundary. Step S4: Calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, convert the actual channel geometry information in the shelf coordinate system to the AMR map coordinate system, generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary, and write the compensation navigation parameters into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. Step S5: Control the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, allow the AMR to enter the target work channel.

[0006] Preferably, in step S4, when writing the compensation navigation parameters into the AMR local cost map, identity tag data including channel number, task number, timestamp, and map version number is synchronously written into the obstacle layer, expansion layer, and restricted area layer, and a channel occupancy lock is set for the AMR; the AMR only calls the compensation navigation parameters to plan the operation path when the obtained map version number is consistent with the version number issued by the collaborative controller and the channel occupancy lock is valid; after the AMR completes the pick-up and drop-off operation and drives out of the target operation channel, the compensation navigation parameters in the obstacle layer, expansion layer, and restricted area layer are automatically cleared.

[0007] Preferably, in step S3, the multi-source detection device further includes a reflector, a radio frequency identification tag, and an ultra-wideband tag installed on the electric mobile shelf or the ground; before calculating the actual stopping position and the actual channel width, the reliability coefficient of each detection sensor in the multi-source detection device is updated based on the consistency error and packet loss rate of the multi-source detection device in a historical number of tasks; when the discrete deviation between the data collected by each detection sensor exceeds the set consistency threshold, the detection data is judged to be abnormal and the AMR is blocked from entering the target operation channel.

[0008] Preferably, in step S4, when converting the actual aisle geometry information in the shelf coordinate system to the AMR map coordinate system, the translation matrix and rotation matrix between the shelf coordinate system and the AMR map coordinate system are determined by at least two non-collinear reference points among the reference block set at the end of the track, the shelf upright reflector, and the ground positioning mark; the coordinate transformation error is checked using the third reference point, and the coordinate transformation recalibration process is triggered when the coordinate transformation error exceeds 0.5°.

[0009] Preferably, in step S5, when the actual physical boundary and the compensated navigation parameters are inconsistent, the following priority levels are applied for processing: if sensor data is missing for a short period of time, the channel boundary data is re-acquired and the compensated navigation parameters are recalculated; if the position deviation exceeds the passage threshold but is less than 50mm, the AMR is controlled to reduce its travel speed and output a remote confirmation request signal; if the angle deviation exceeds 1.0° or the channel width deviation exceeds 30mm, the AMR is prohibited from entering and the electric mobile shelf is controlled to reopen the channel; when the number of consecutive comparison failures reaches 3, the target operation channel is frozen and a maintenance work order is generated.

[0010] Preferably, after step S4, the method further includes: step S6, storing the actual stopping position and position deviation after each movement of the electric mobile shelf, constructing a historical stopping deviation time series model for the corresponding shelf; calculating the deviation accumulation rate of the historical stopping deviation time series model within the sliding time window; when the deviation accumulation rate continuously exceeds 5mm / h, outputting a warning signal for the electric mobile shelf drive mechanism or track, and increasing the safety margin threshold for subsequent picking and placing tasks of the electric mobile shelf.

[0011] Preferably, in step S4, the generated compensation navigation parameters also include a deceleration area and a stop area limited to the target work channel; after the AMR enters the target work channel, the AMR is controlled to reduce its speed in the deceleration area and complete the docking and retrieval operation with the target storage location in the stop area; when the target work channel fails or the electric mobile rack receives a new movement instruction, the permission status of the target work channel is canceled and the deceleration area and stop area are deleted.

[0012] Preferably, in step S4, the compensation navigation parameters also include a guide sector set at the front end of the target operation channel entrance; when the position deviation is greater than 20mm or the angle deviation is greater than 0.8°, the angle of the guide sector in the AMR map coordinate system is expanded, and the expansion coefficient of the cost grid in the guide sector is increased, so that the AMR can adjust its attitude with the increased turning radius and drive into the center line of the target operation channel.

[0013] Preferably, in step S5, when the AMR scans the entrance contour of the target operation channel using the vehicle-mounted LiDAR, it includes the following sub-steps: Step S51, obtain the actual deformation contour of the high-level uprights of the shelves on both sides of the target operation channel, and extract the cargo over-edge dimensions that exceed the edge of the shelf base; when the actual deformation contour or cargo over-edge dimensions intrude into the driving safety space of the AMR, determine that the actual physical boundary is inconsistent with the compensation navigation parameters, and prevent the AMR from driving in.

[0014] An AMR navigation compensation system for parking deviation of an electric mobile shelving unit includes: The location acquisition module is used to obtain the theoretical stopping position of the target shelf and the target storage location; The rack drive module is used to drive the electric mobile rack to move and form the target operation channel; The data acquisition and fusion module is used to collect rack position data and aisle boundary data in stages through multi-source detection devices. It dynamically adjusts the fusion weight of the multi-source detection devices according to the operation stage of the electric mobile rack. During the start-up and movement stages, the weight of the encoder and drive current sensor is increased. During the approach to stop stage, the weight of the locking pin status sensor is increased. During the post-stop verification stage, the weight of the lidar and visual marker is increased. The actual stop position, actual aisle width and aisle boundary are calculated. The deviation calculation and map update module is used to calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, the actual channel geometry information under the shelf coordinate system is converted to the AMR map coordinate system to generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary. The compensation navigation parameters are then written into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. The access control module controls the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, the AMR is permitted to enter the target work channel.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. The control system obtains the actual stopping position and theoretical preset position of the target structural component, extracts the position offset amplitude, angular tilt angle and channel space width difference, and converts the above physical geometric variables into interface update parameters of the vehicle local map. By writing the corrected channel geometric data into the obstacle layer, expansion layer and restricted area of ​​the local cost map in real time, the moving vehicle completes the driving boundary reconstruction before entering the dynamic channel. This spatial parameter transmission and interface reconstruction mechanism eliminates the cumulative positioning deviation caused by mechanical drive inertia, track gap and wheel-rail wear, and avoids the risk of physical scratches during the vehicle's entry into the channel and docking at the storage location.

[0016] 2. During the structural component movement, approach-stop phase, and post-stop verification phase, the system adaptively matches the fusion weights of encoder pulses, limit switch signals, locking pin status, laser contour data, and visual marker data. When the consistency deviation of multi-source sensor data exceeds a preset threshold, the generation of passage permits is automatically blocked. The actual channel space geometric parameters are only output after the data consistency verification is passed. This dynamic weight allocation and consistency verification mechanism based on the motion phase isolates the measurement distortion caused by physical obstruction, optical interference, or mechanical slippage of a single sensor, ensuring that the channel space characteristics input to the navigation compensation network have solid physical authenticity.

[0017] 3. When the moving vehicle arrives at the entrance of the passage, the on-board sensing equipment is used to conduct a field scan of the actual physical edge of the passage. The compensation map issued by the main control system is compared with the geometric contour obtained by the field scan. When the field scan data and the compensation map data are consistent, an entry permit is generated. When the verification is inconsistent, the sensor re-sampling and fusion, low-speed trial passage and passage freezing maintenance process are initiated according to the degree of deviation. This secondary verification and graded response mechanism of the entrance supplements and eliminates physical space anomalies that cannot be perceived by the main control benchmark, such as mechanical deformation of structural frame, goods exceeding the boundary or temporary obstacles, thus building a double physical safety defense barrier. Attached Figure Description

[0018] Figure 1 This is a flowchart of the channel spatial deviation calculation and AMR map update process of the present invention; Figure 2 This is a block diagram of the collaborative control system and multi-source detection equipment module of the present invention.

[0019] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0021] A method for compensating for parking deviation of an electric mobile shelving unit using AMR navigation includes the following steps: Step S1: Obtain the theoretical stopping positions of the target shelf and the target storage location; Step S2: Drive the electric mobile shelf to move and form the target operation channel; Step S3: Collect rack position data and aisle boundary data in stages using a multi-source detection device. Dynamically adjust the fusion weight of the multi-source detection device according to the operation stage of the electric mobile rack. Increase the weight of encoder and drive current sensor in the start-up and movement stages, increase the weight of locking pin status sensor in the near-stop stage, and increase the weight of lidar and visual marker in the post-stop verification stage. Calculate the actual stop position, actual aisle width, and aisle boundary. Step S4: Calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, convert the actual channel geometry information in the shelf coordinate system to the AMR map coordinate system, generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary, and write the compensation navigation parameters into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. Step S5: Control the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, allow the AMR to enter the target work channel.

[0022] Preferably, in step S4, when writing the compensation navigation parameters into the AMR local cost map, identity tag data including channel number, task number, timestamp, and map version number is synchronously written into the obstacle layer, expansion layer, and restricted area layer, and a channel occupancy lock is set for the AMR; the AMR only calls the compensation navigation parameters to plan the operation path when the obtained map version number is consistent with the version number issued by the collaborative controller and the channel occupancy lock is valid; after the AMR completes the pick-up and drop-off operation and drives out of the target operation channel, the compensation navigation parameters in the obstacle layer, expansion layer, and restricted area layer are automatically cleared.

[0023] Preferably, in step S3, the multi-source detection device further includes a reflector, a radio frequency identification tag, and an ultra-wideband tag installed on the electric mobile shelf or the ground; before calculating the actual stopping position and the actual channel width, the reliability coefficient of each detection sensor in the multi-source detection device is updated based on the consistency error and packet loss rate of the multi-source detection device in a historical number of tasks; when the discrete deviation between the data collected by each detection sensor exceeds the set consistency threshold, the detection data is judged to be abnormal and the AMR is blocked from entering the target operation channel.

[0024] Preferably, in step S4, when converting the actual aisle geometry information in the shelf coordinate system to the AMR map coordinate system, the translation matrix and rotation matrix between the shelf coordinate system and the AMR map coordinate system are determined by at least two non-collinear reference points among the reference block set at the end of the track, the shelf upright reflector, and the ground positioning mark; the coordinate transformation error is checked using the third reference point, and the coordinate transformation recalibration process is triggered when the coordinate transformation error exceeds 0.5°.

[0025] Preferably, in step S5, when the actual physical boundary and the compensated navigation parameters are inconsistent, the following priority levels are applied for processing: if sensor data is missing for a short period of time, the channel boundary data is re-acquired and the compensated navigation parameters are recalculated; if the position deviation exceeds the passage threshold but is less than 50mm, the AMR is controlled to reduce its travel speed and output a remote confirmation request signal; if the angle deviation exceeds 1.0° or the channel width deviation exceeds 30mm, the AMR is prohibited from entering and the electric mobile shelf is controlled to reopen the channel; when the number of consecutive comparison failures reaches 3, the target operation channel is frozen and a maintenance work order is generated.

[0026] Preferably, after step S4, the method further includes: step S6, storing the actual stopping position and position deviation after each movement of the electric mobile shelf, constructing a historical stopping deviation time series model for the corresponding shelf; calculating the deviation accumulation rate of the historical stopping deviation time series model within the sliding time window; when the deviation accumulation rate continuously exceeds 5mm / h, outputting a warning signal for the electric mobile shelf drive mechanism or track, and increasing the safety margin threshold for subsequent picking and placing tasks of the electric mobile shelf.

[0027] Preferably, in step S4, the generated compensation navigation parameters also include a deceleration area and a stop area limited to the target work channel; after the AMR enters the target work channel, the AMR is controlled to reduce its speed in the deceleration area and complete the docking and retrieval operation with the target storage location in the stop area; when the target work channel fails or the electric mobile rack receives a new movement instruction, the permission status of the target work channel is canceled and the deceleration area and stop area are deleted.

[0028] Preferably, in step S4, the compensation navigation parameters also include a guide sector set at the front end of the target operation channel entrance; when the position deviation is greater than 20mm or the angle deviation is greater than 0.8°, the angle of the guide sector in the AMR map coordinate system is expanded, and the expansion coefficient of the cost grid in the guide sector is increased, so that the AMR can adjust its attitude with the increased turning radius and drive into the center line of the target operation channel.

[0029] Preferably, in step S5, when the AMR scans the entrance contour of the target operation channel using the vehicle-mounted LiDAR, it includes the following sub-steps: Step S51, obtain the actual deformation contour of the high-level uprights of the shelves on both sides of the target operation channel, and extract the cargo over-edge dimensions that exceed the edge of the shelf base; when the actual deformation contour or cargo over-edge dimensions intrude into the driving safety space of the AMR, determine that the actual physical boundary is inconsistent with the compensation navigation parameters, and prevent the AMR from driving in.

[0030] An AMR navigation compensation system for parking deviation of an electric mobile shelving unit includes: The location acquisition module is used to obtain the theoretical stopping position of the target shelf and the target storage location; The rack drive module is used to drive the electric mobile rack to move and form the target operation channel; The data acquisition and fusion module is used to collect rack position data and aisle boundary data in stages through multi-source detection devices. It dynamically adjusts the fusion weight of the multi-source detection devices according to the operation stage of the electric mobile rack. During the start-up and movement stages, the weight of the encoder and drive current sensor is increased. During the approach to stop stage, the weight of the locking pin status sensor is increased. During the post-stop verification stage, the weight of the lidar and visual marker is increased. The actual stop position, actual aisle width and aisle boundary are calculated. The deviation calculation and map update module is used to calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, the actual channel geometry information under the shelf coordinate system is converted to the AMR map coordinate system to generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary. The compensation navigation parameters are then written into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. The access control module controls the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, the AMR is permitted to enter the target work channel.

[0031] Example 1: This example combines Figures 1 to 2 This paper explains the AMR navigation compensation method and system for parking deviation of electric mobile shelving, such as... Figure 1 As shown, the process first executes step S1: obtaining the theoretical stopping position, acquiring the theoretical stopping position of the target shelf and the target storage location; then executes step S2: forming the target operation channel, driving the electric mobile shelf to move and form the target operation channel; subsequently, the shelf stopping position and channel boundary data are collected through a multi-source detection device, and step S3: calculating the actual stopping position and boundary, collecting data in stages and dynamically adjusting the fusion weight to calculate the actual channel width and channel boundary; next, step S4: generating and writing compensation navigation parameters, calculating the deviation and converting it to the AMR map coordinate system to generate compensation navigation parameters and writing them into the cost map, writing the generated compensation navigation parameters into the AMR local cost map containing the obstacle layer, expansion layer and restricted layer; finally, step S5: comparing and permitting AMR entry, using onboard sensors to scan the channel entrance contour, scanning the entrance contour of the target operation channel, and permitting AMR entry when the physical boundary matches.

[0032] like Figure 2As shown, the electric mobile rack is equipped with a rack drive module, encoder, drive current sensor, and locking pin status sensor, and transmits rack position data to the collaborative controller. The multi-source detection device and physical environment are equipped with visual markers and reflectors, RFID tags and ultra-wideband tags, as well as reference blocks and ground positioning marks at the end of the track, and transmit channel boundary data to the collaborative controller. The collaborative controller integrates a position acquisition module, a data acquisition and fusion module, a deviation calculation and map update module, and an access control module. The collaborative controller processes the received data and sends compensation navigation parameters and map version numbers to the AMR. The AMR is equipped with vehicle-mounted sensors and vehicle-mounted LiDAR, as well as a local cost map of the AMR including obstacle layers, expansion layers, and restricted areas. The AMR uses vehicle-mounted sensors and vehicle-mounted LiDAR to scan the target operation channel entrance contour of the multi-source detection device and physical environment.

[0033] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for compensating for parking deviation of an electric mobile shelving unit using AMR navigation, characterized in that, Includes the following steps: Step S1: Obtain the theoretical stopping positions of the target shelf and the target storage location; Step S2: Drive the electric mobile shelf to move and form the target operation channel; Step S3: Collect rack position data and aisle boundary data in stages using a multi-source detection device. Dynamically adjust the fusion weight of the multi-source detection device according to the operation stage of the electric mobile rack. Increase the weight of encoder and drive current sensor in the start-up and movement stages, increase the weight of locking pin status sensor in the near-stop stage, and increase the weight of lidar and visual marker in the post-stop verification stage. Calculate the actual stop position, actual aisle width, and aisle boundary. Step S4: Calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, convert the actual channel geometry information in the shelf coordinate system to the AMR map coordinate system, generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary, and write the compensation navigation parameters into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. Step S5: Control the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, allow the AMR to enter the target work channel.

2. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S4, when writing the compensated navigation parameters into the AMR local cost map, identity tag data containing channel number, task number, timestamp and map version number is synchronously written into the obstacle layer, expansion layer and restricted layer, and a channel occupancy lock is set for the AMR. AMR only calls the compensation navigation parameters to plan the operation path when the obtained map version number is consistent with the version number issued by the collaborative controller and the channel occupancy lock is valid. After the AMR completes the pick-up and drop-off operation and drives out of the target operation channel, it automatically clears the compensation navigation parameters in the obstacle layer, expansion layer, and restricted area layer.

3. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S3, the multi-source detection device also includes reflectors, radio frequency identification tags, and ultra-wideband tags installed on the electric mobile shelf or the ground; before calculating the actual stopping position and the actual channel width, the reliability coefficients of each detection sensor in the multi-source detection device are updated based on the consistency error and packet loss rate of the multi-source detection device in a historical number of tasks; when the discrete deviation between the data collected by each detection sensor exceeds the set consistency threshold, the detection data is judged to be abnormal and the AMR is blocked from entering the target operation channel.

4. A method for compensating for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S4, when converting the actual aisle geometry information in the shelf coordinate system to the AMR map coordinate system, the translation and rotation matrices between the shelf coordinate system and the AMR map coordinate system are determined by using at least two non-collinear reference points among the reference blocks set at the ends of the tracks, the reflectors on the shelf uprights, and the ground positioning marks. The coordinate transformation error is checked using the third reference point, and the coordinate transformation recalibration process is triggered when the coordinate transformation error exceeds 0.5°.

5. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S5, when the actual physical boundary does not match the compensated navigation parameters, the following priority levels are applied: If sensor data is temporarily missing, the channel boundary data is re-acquired and the compensated navigation parameters are recalculated; if the position deviation exceeds the passage threshold but is less than 50mm, the AMR is controlled to reduce its travel speed and a remote confirmation request signal is output; if the angle deviation exceeds 1.0° or the channel width deviation exceeds 30mm, the AMR is prohibited from entering and the electric mobile shelf is controlled to reopen the channel; if the number of consecutive comparison failures reaches 3, the target operation channel is frozen and a maintenance work order is generated.

6. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, After step S4, the following steps are also included: Step S6, storing the actual stopping position and position deviation after each movement of the electric mobile shelf, constructing a historical stopping deviation time series model for the corresponding shelf; calculating the deviation accumulation rate of the historical stopping deviation time series model within the sliding time window; when the deviation accumulation rate continuously exceeds 5mm / h, outputting a warning signal for the electric mobile shelf drive mechanism or track, and increasing the safety margin threshold for subsequent picking and placing tasks of the electric mobile shelf.

7. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S4, the generated compensation navigation parameters also include a deceleration area and a stopping area limited to the target work channel; after the AMR enters the target work channel, the AMR is controlled to reduce its speed in the deceleration area and complete the docking and retrieval operation with the target storage location in the stopping area. When the target work lane becomes unavailable or the electric mobile rack receives a new movement instruction, cancel the permission status of the target work lane and delete the deceleration zone and stop zone.

8. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S4, the compensation navigation parameters also include a guide sector set at the front end of the target operation channel entrance; when the position deviation is greater than 20mm or the angle deviation is greater than 0.8°, the angle of the guide sector in the AMR map coordinate system is expanded, and the expansion coefficient of the cost grid in the guide sector is increased, so that the AMR can adjust its attitude with the increased turning radius and drive into the center line of the target operation channel.

9. The AMR navigation compensation method for parking deviation of an electric mobile shelving unit according to claim 1, characterized in that, In step S5, when the AMR scans the entrance contour of the target operation channel using the vehicle-mounted LiDAR, it includes the following sub-steps: Step S51, obtain the actual deformation contour of the high-level uprights of the shelves on both sides of the target operation channel, and extract the cargo over-edge dimensions that exceed the edge of the shelf base; when the actual deformation contour or cargo over-edge dimensions intrude into the driving safety space of the AMR, determine that the actual physical boundary is inconsistent with the compensation navigation parameters, and prevent the AMR from driving in.

10. An AMR navigation compensation system for parking deviation of an electric mobile shelving unit, used to implement the AMR navigation compensation method for parking deviation of an electric mobile shelving unit as described in any one of claims 1 to 9, characterized in that, include: The location acquisition module is used to obtain the theoretical stopping position of the target shelf and the target storage location; The rack drive module is used to drive the electric mobile rack to move and form the target operation channel; The data acquisition and fusion module is used to collect rack position data and aisle boundary data in stages through multi-source detection devices. It dynamically adjusts the fusion weight of the multi-source detection devices according to the operation stage of the electric mobile rack. During the start-up and movement stages, the weight of the encoder and drive current sensor is increased. During the approach to stop stage, the weight of the locking pin status sensor is increased. During the post-stop verification stage, the weight of the lidar and visual marker is increased. The actual stop position, actual aisle width and aisle boundary are calculated. The deviation calculation and map update module is used to calculate the positional deviation, angle deviation, and channel width deviation between the theoretical stopping position and the actual stopping position. When the positional deviation, angle deviation, and channel width deviation are within a certain range, the actual channel geometry information under the shelf coordinate system is converted to the AMR map coordinate system to generate compensation navigation parameters including the entrance point, channel centerline, and restricted boundary. The compensation navigation parameters are then written into the obstacle layer, expansion layer, and restricted layer of the AMR local cost map. The access control module controls the AMR to scan the entrance contour of the target work channel using onboard sensors. When the actual physical boundary matches the compensated navigation parameters, the AMR is permitted to enter the target work channel.

Images