Vehicle handling system

Abstract

This application discloses a vehicle handling system. The system's dispatch center can generate handling tasks containing handling paths; filter all target AGVs for executing the handling tasks; select master and slave AGVs when the number of target AGVs is not one; receive the latest first location information uploaded by the master AGV; and generate a first motion target for the vehicle to be handled based on the latest first location information and the handling path. The master AGV can receive the latest first motion target issued by the dispatch center and generate a first posture adjustment state for each slave AGV based on the latest first motion target; issue the corresponding first posture adjustment state to each slave AGV; and generate and upload the first location information of the vehicle to be handled when the vehicle's progress towards the first motion target reaches a preset progress threshold. Therefore, this application can improve dynamic adaptability and fault tolerance through a hierarchical control mode of dispatch center coordination and master-slave AGV collaborative handling.

Description

Technical Field

[0001] This application relates to the field of robotics, and more specifically, to a vehicle handling system. Background Technology

[0002] In parking lots, users driving their own cars to find parking spaces is not only time-consuming and laborious, but also prone to scratches due to space constraints or lack of driving experience, seriously affecting parking efficiency and vehicle safety. To solve these problems, existing technology introduces AGV parking robot solutions. Multiple AGVs work together to move vehicles parked at designated entrances to available parking spaces using mechanical support or towing, significantly reducing the difficulty of operation for users.

[0003] However, existing AGV collaborative handling technology, designed for a single vehicle handling task, generates only one scheduling strategy at the coordination center. Subsequent AGVs performing the task follow this strategy without further adjustments. But vehicle handling is not simply path following; it must address dynamic scenarios such as changes in ground flatness and new path failures. A statically generated strategy cannot adapt to these unforeseen circumstances in real time. If situations arise outside the strategy's preset parameters, AGV malfunctions are highly likely, impacting handling safety and efficiency. Therefore, optimizing AGV collaborative handling technology and improving its fault tolerance has become a crucial issue for those skilled in the art. Summary of the Invention

[0004] In view of this, this application provides a vehicle handling system to address the shortcomings of poor fault tolerance in existing AGV collaborative handling technology.

[0005] To achieve the above objectives, the following solution is proposed:

[0006] A vehicle handling system includes a dispatch center and multiple AGVs;

[0007] The dispatch center determines the basic information of the vehicle to be transported and generates a transport task containing the transport path; it selects all target AGVs from the various AGVs to perform the transport task; when the number of target AGVs is not 1, it selects a master AGV from the target AGVs and uses the other target AGVs as slave AGVs; it receives the latest first location information uploaded by the master AGV and generates a first motion target for the vehicle to be transported based on the latest first location information and the transport path.

[0008] The master AGV receives the latest first motion target issued by the scheduling center, and generates a first posture adjustment state for each slave AGV based on the latest first motion target; it issues a corresponding first posture adjustment state to each slave AGV so that each slave AGV and the master AGV can work together to drive the vehicle to be transported to achieve the first motion target; when the progress of the vehicle to be transported toward the first motion target reaches a preset progress threshold, it generates and uploads the first position information of the vehicle to be transported.

[0009] Optional, also includes:

[0010] When the number of target AGVs is 1, the dispatch center receives the latest second location information uploaded by the target AGV, and generates the second motion target of the vehicle to be transported based on the latest second location information and the transport path.

[0011] The target AGV receives the latest second motion target issued by the scheduling center; based on the latest second motion target, it calculates a second posture adjustment state and adjusts the motion posture of the target AGV to the second posture adjustment state to drive the vehicle to be transported to achieve the second motion target; when the progress of the vehicle to be transported toward the second motion target reaches a preset progress threshold, it generates and uploads the latest second position information of the vehicle to be transported.

[0012] Optionally, the scheduling center may also include an AGV screening module;

[0013] The AGV screening module determines the number of AGVs based on the handling task; calculates a task score based on the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree, and health status of each AGV; and selects target AGVs that match the number of AGVs from the AGVs according to the corresponding task score.

[0014] Optionally, the AGV screening module includes a scoring calculation unit;

[0015] The scoring calculation unit predicts the energy consumption cost and time cost of the corresponding AGV to reach the handling task receiving point based on the current position of each AGV, and calculates the energy consumption cost and time cost corresponding to the same AGV in a weighted manner to obtain the task cost of the corresponding AGV; the ratio between the remaining number of tasks of each AGV and the historical average daily load of the AGV is used as the load saturation coefficient; the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree and health status corresponding to the same AGV are weighted to obtain the task score of the corresponding AGV.

[0016] Optionally, the scheduling center includes a path generation unit;

[0017] The path generation unit acquires a target location map, analyzes the target location map, and determines the distribution of feasible road segments; based on the distribution of feasible road segments in the target location map, it plans a transportation route; it analyzes the transportation route, predicts road segment occupancy data, and updates the distribution of feasible road segments in the target location map based on the road segment occupancy data.

[0018] Optionally, the path generation unit includes a road segment occupancy prediction component;

[0019] The road occupancy prediction component predicts the occupancy time of each road segment based on all traffic segments included in the transport path; according to the basic information and the size of each traffic segment in the target location map, it evaluates whether the corresponding traffic segment can still allow other vehicles to pass after the vehicle to be transported has passed, and obtains the evaluation result of the corresponding traffic segment; in the target location map, it marks the corresponding occupancy time at the matching point of the traffic segment whose evaluation result is not allowed.

[0020] Optionally, the path generation unit further includes a fault occupancy update component;

[0021] The fault occupancy update component responds to the road segment occupancy status reported by the AGV that has stopped operating due to a fault, and updates the target location map based on the road segment occupancy status to distribute feasible road segments.

[0022] Optionally, the scheduling center may further include a moving target generation unit;

[0023] The motion target generation unit determines a theoretical position matching the latest first position information from the transport path, selects the next position of the theoretical position from the transport path as the first motion endpoint, analyzes the constraint factors corresponding to the arrival of the first motion endpoint from the latest first position information, determines the first motion linear velocity and the first motion angular velocity, and generates a first motion target including the first motion endpoint, the first motion linear velocity, and the first motion angular velocity.

[0024] Optionally, the main AGV further includes an attitude adjustment generation unit;

[0025] The attitude adjustment generation unit determines the coordinate position of each target AGV relative to the vehicle center; based on the coordinate position of each target AGV and the first linear velocity and first angular velocity of motion, it generates the forward velocity component along the vehicle's forward direction and the left and right velocity components along the vehicle's left and right directions for the corresponding target AGV; based on the forward velocity component and the left and right velocity components of each target AGV, it calculates the target running speed and target steering angle of the corresponding target AGV; based on the target running speed and target steering angle of each target AGV, it generates the first attitude adjustment state of the corresponding target AGV.

[0026] Optionally, the attitude adjustment generation unit includes an attitude decomposition component;

[0027] The attitude decomposition component determines the total attitude deviation of each target AGV by combining the target running speed, target turning angle, and current motion attitude of each target AGV. It then adds the corresponding total attitude deviation to the current motion attitude of each target AGV in the same proportion to obtain the first attitude adjustment state of the corresponding target AGV, thereby updating the current motion attitude of each target AGV. The process continues until the current motion attitude of each target AGV satisfies the corresponding target running speed and target turning angle.

[0028] As can be seen from the above technical solution, the vehicle handling system provided in this application can include a dispatch center and multiple AGVs. The dispatch center determines the basic information of the vehicle to be handled and generates a handling task containing the handling path. From the various AGVs, it selects all target AGVs to perform the handling task. When the number of target AGVs is not 1, it selects a master AGV from the target AGVs and uses the other target AGVs as slave AGVs. It receives the latest first position information uploaded by the master AGV and generates a first motion target of the vehicle to be handled based on the latest first position information and the handling path. Based on this, each handling task of the dispatch center in this application does not correspond to a static and unchanging dispatch strategy, but is dynamically generated based on the position information uploaded by the master AGV in real time. Therefore, the first motion target generated by the dispatch center can respond in real time to the displacement caused by slight slippage of the AGV and uneven ground during the handling process, avoiding handling deviations caused by the initial preset strategy being out of sync with the actual working conditions, ensuring that the motion target always fits the current actual position, and improving the path following accuracy in complex scenarios. The master AGV receives the latest first motion target from the scheduling center and, based on this target, generates a first posture adjustment state for each slave AGV. It then issues the corresponding first posture adjustment state to each slave AGV, enabling coordinated action between the slave AGVs and the master AGV to drive the vehicle to be transported to achieve the first motion target. When the vehicle's progress towards the first motion target reaches a preset progress threshold, the master AGV generates and uploads the vehicle's first position information. This method allows the master AGV to independently receive scheduling instructions, avoiding asynchronous actions caused by varying computing capabilities or instruction interpretations among multiple AGVs during transport tasks. This effectively reduces vehicle tilting and uneven load caused by uncoordinated AGV movements, ensuring the stability of the entire vehicle transport. The master AGV only uploads position information and triggers the next round of motion target generation when the vehicle reaches the preset progress threshold for the first motion target, forming a closed-loop control system for target issuance, action execution, progress feedback, and new target generation. This phased management approach avoids process loss due to issuing full-path instructions all at once, ensuring that each transport action is completed within a controllable range. This facilitates timely detection and correction of local deviations, improving the reliability and traceability of the overall transport task and increasing its overall fault tolerance. Furthermore, this application generates motion targets in stages by the scheduling center, and the main AGV calculates the posture targets of different AGVs based on these motion targets, reducing the computational burden on the scheduling center. The scheduling center no longer needs to handle the complex calculations of posture allocation and collaborative logic for all AGVs along the entire path at once; it only needs to focus on the core task of generating phased motion targets, significantly reducing data processing volume and real-time computation load, and avoiding response delays caused by parallel computing of multiple tasks.Meanwhile, the master AGV, as the core of local collaboration, can flexibly calculate the posture target based on the current actual handling scenario. Compared with remote command issuance from the scheduling center, this reduces the time spent on command transmission links and adaptation, allowing AGV posture adjustments to better fit dynamic working conditions, improving collaborative response speed, and further reducing the overall operational pressure on the scheduling center, ensuring system stability during multi-task concurrency. Therefore, this application, through a hierarchical control mode of scheduling center coordination and master-slave AGV collaborative handling, improves dynamic adaptability and fault tolerance, reduces AGV asynchrony caused by transmission delays or interpretation differences, and lowers handling risks. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0030] Figure 1 This is a system architecture diagram of a vehicle handling system disclosed in an embodiment of this application;

[0031] Figure 1 The correspondence between component identification and figure reference numerals can be seen as follows:

[0032] Dispatch Center 10, AGV 20, Master AGV 200, Slave AGV 201. Detailed Implementation

[0033] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0034] The vehicle handling system of this application will now be described in detail with reference to the accompanying drawings. It should be noted that the orientation of the structures shown in the drawings is set for ease of understanding and does not limit the orientation of the embodiments of this disclosure in actual implementation. Furthermore, the shape and size of the structure, whether as a whole or in part, shown in the drawings are not limited to the actual shape and size.

[0035] Next, combine Figure 1 This application provides a detailed description of the vehicle handling system.

[0036] See Figure 1It can be seen that the dispatch center of this application may include a dispatch center 10 and multiple AGVs 20.

[0037] The dispatch center 10 can interact with each AGV 20.

[0038] The dispatch center 10 can be a smart device configured on the network side. For example, the dispatch center 10 can be a cloud server or a mobile terminal.

[0039] The AGV20 can be a mobile robot that includes an adaptive lifting mechanism for transporting vehicles.

[0040] Different AGV20 models can be the same or different.

[0041] During the application process, the dispatch center 10 can use a smart camera to determine the basic information of the vehicle to be transported and generate a transport task containing the transport path; from each AGV 20, it selects all target AGVs to perform the transport task; when the number of target AGVs is not 1, it selects a master AGV 200 from each target AGV and uses the other target AGVs as slave AGVs 201; it receives the latest first location information uploaded by the master AGV 200, and based on the latest first location information and the transport path, it generates the first motion target of the vehicle to be transported.

[0042] The master AGV 200 receives the latest first motion target issued by the scheduling center 10, and generates a first posture adjustment state for each slave AGV 201 based on the latest first motion target; it issues the corresponding first posture adjustment state to each slave AGV 201 so that each slave AGV 201 and the master AGV 200 coordinate their actions to drive the vehicle to be transported to achieve the first motion target; when the progress of the vehicle to be transported toward the first motion target reaches a preset progress threshold, it generates and uploads the first position information of the vehicle to be transported.

[0043] The basic information may include the vehicle type, dimensions, wheelbase, and track width of the vehicle to be moved, as well as the starting point of the move.

[0044] The dispatch center 10 can determine the destination of the vehicle to be moved based on the availability of parking spaces.

[0045] The dispatch center 10 can generate the transport path of the vehicle to be transported based on the transport start point and transport end point, generate the transport task based on the transport path and basic information, and determine the number of AGVs matched with the vehicle to be transported based on the transport-related data.

[0046] The master AGV200 can generate different first posture adjustment states for the slave AGV201s based on the latest first motion target, so that each slave AGV201 drives the vehicle to be transported to achieve the first motion target. The movement status of the vehicle to be transported can be compared with the execution progress of the first motion target in real time. When the execution progress of the movement status of the vehicle to be transported reaches the preset progress threshold of the first motion target, the real-time position of the vehicle to be transported is uploaded as the first position information.

[0047] The progress threshold can be determined based on the remaining resources of scheduling center 10.

[0048] For example, if the remaining resources of the scheduling center 10 are small, it means that it will take longer to calculate the moving target. In this case, the progress threshold can be set smaller to reserve sufficient calculation time for the scheduling center 10. Conversely, if the remaining resources of the scheduling center 10 are sufficient and the calculation efficiency is higher, the progress threshold can be appropriately increased to improve the overall transportation rhythm.

[0049] When the dispatch center 10 receives multiple handling tasks, it can calculate the priority of each handling task.

[0050] The priority of the vehicle retrieval task is higher than the priority of the vehicle storage task.

[0051] The longer the waiting time, the higher the priority of the transfer task.

[0052] As can be seen from the above technical solution, the vehicle handling system provided in this application can include a dispatch center 10 and multiple AGVs 20. The dispatch center determines the basic information of the vehicle to be handled and generates a handling task containing the handling path. From each AGV 20, it selects all target AGVs to perform the handling task. When the number of target AGVs is not 1, it selects a master AGV 200 from the target AGVs and uses the other target AGVs as slave AGVs 201. It receives the latest first position information uploaded by the master AGV 200 and generates a first motion target of the vehicle to be handled based on the latest first position information and the handling path. Based on this, each handling task of the dispatch center in this application does not correspond to a static and unchanging dispatch strategy, but is dynamically generated based on the position information uploaded by the master AGV 200 in real time. Therefore, the first motion target generated by the dispatch center can respond in real time to the displacement caused by slight slippage of AGVs and uneven ground during the handling process, avoiding handling deviations caused by the initial preset strategy being out of sync with the actual working conditions, ensuring that the motion target always fits the current actual position, and improving the path following accuracy in complex scenarios. The master AGV 200 receives the latest first motion target from the scheduling center and, based on this target, generates a first posture adjustment state for each slave AGV 201. It then sends the corresponding first posture adjustment state to each slave AGV 201, enabling coordinated action between the slave AGVs and the master AGV 200 to drive the vehicle to be transported to achieve the first motion target. When the vehicle's progress towards the first motion target reaches a preset progress threshold, the master AGV 200 generates and uploads the vehicle's first position information. This method allows the master AGV 200 to independently receive scheduling instructions, avoiding asynchronous actions caused by varying computing capabilities or instruction interpretation among multiple AGVs performing transport tasks. This effectively reduces vehicle tilting and uneven load caused by uncoordinated AGV movements, ensuring the stability of the entire vehicle transport. The master AGV 200 only uploads position information and triggers the next round of motion target generation when the vehicle reaches the preset progress threshold for the first motion target, forming a closed-loop control system for target issuance, action execution, progress feedback, and new target generation. This phased management approach avoids process loss due to issuing full-path instructions all at once, ensuring that each transport action is completed within a controllable range. This facilitates timely detection and correction of local deviations, improving the reliability and traceability of the overall transport task and increasing its overall fault tolerance. Furthermore, this application uses a phased generation of motion targets by the scheduling center, with the main AGV200 calculating the attitude targets of different AGV20s based on these motion targets, reducing the computational burden on the scheduling center.The scheduling center 10 does not need to handle complex calculations such as attitude allocation and collaborative logic for all AGV20s along the entire path at once. It only needs to focus on the core task of generating phased motion targets, significantly reducing data processing volume and real-time computing load, and avoiding response delays caused by multi-task parallel computing. Meanwhile, the master AGV200, as the local collaborative core, can flexibly calculate attitude targets based on the current actual handling scenario. Compared to remote command issuance from the scheduling center, this reduces command transmission links and adaptation time, allowing AGV20 attitude adjustments to better fit dynamic working conditions, improving collaborative response speed, and further reducing the overall operational pressure on the scheduling center, ensuring system stability during multi-task concurrency. Therefore, this application, through a hierarchical control mode of scheduling center coordination and master-slave AGV collaborative handling, improves dynamic adaptability and fault tolerance, reduces AGV20 asynchrony caused by transmission delays or interpretation differences, and lowers handling risks.

[0053] In some embodiments of this application, it is considered that only one AGV20 is needed to complete the transportation task for vehicles of a specific model.

[0054] At this point, the number of target AGVs can be 1.

[0055] When the number of target AGVs is 1, the dispatch center 10 can directly receive the latest second location information uploaded by the target AGV, and generate the second motion target of the vehicle to be transported based on the latest second location information and the transport path.

[0056] The target AGV can receive the latest second motion target issued by the scheduling center 10; based on the latest second motion target, it calculates the second posture adjustment state and adjusts the motion posture of the target AGV to the second posture adjustment state to drive the vehicle to be transported to achieve the second motion target; when the execution progress of the vehicle to be transported to the second motion target reaches a preset progress threshold, the latest second position information of the vehicle to be transported is generated and uploaded.

[0057] It should be noted that the first location information refers to the vehicle location information uploaded by the main AGV200, while the second location information refers to the vehicle location information uploaded by the target AGV, which has a quantity of 1.

[0058] The first and second location information belong to the location information of different handling tasks, and the specific locations they match can be the same or different.

[0059] The first motion objective is the phased motion expectation generated by the dispatch center 10 for multiple target AGVs to collaboratively complete the material handling task;

[0060] The second motion objective is the phased motion expectation generated by the dispatch center 10 for a target AGV to complete the transport task.

[0061] The first attitude adjustment state refers to the attitude adjustment parameters calculated by the main AGV200, while the second position information refers to the attitude adjustment parameters calculated by the target AGV, which has a quantity of 1.

[0062] As can be seen from the above technical solution, this embodiment provides a collaborative method between the scheduling center 10 and the target AGV when a single AGV 20 performs a transport task. Through this method, the transport task can still be implemented in stages, allowing this application to be applied to single AGV transport scenarios and improving its applicability.

[0063] In some embodiments of this application, the scheduling center 10 may also include an AGV screening module.

[0064] The AGV screening module can determine the number of AGVs based on the transport task; calculate the task score based on the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree, and health status of each AGV20; and select the target AGV that matches the number of AGVs from each AGV20 according to the size of the corresponding task score.

[0065] Specifically, a matching table containing the correspondence between different vehicle types and the quantity of goods transported can be pre-built.

[0066] Based on the relevant information about the transport, the type of vehicle to be transported can be determined, and the number of AGVs corresponding to that type can be selected from the matching table.

[0067] The state score of each state parameter can be determined based on the state parameters of each AGV20.

[0068] The task score is obtained by weighting the scores of each state.

[0069] Among them, the status parameters can be idle state, current location, task cost, load saturation coefficient, battery retention status, vehicle model matching degree, and health status.

[0070] The idle status of each AGV20 can represent the length of the remaining task queue of the corresponding AGV20;

[0071] The longer the remaining task queue, the smaller the corresponding status score.

[0072] The current position of each AGV20 can represent the location of the corresponding AGV20, and can be used to calculate the distance between each AGV20 and the vehicle to be transported;

[0073] The longer the distance, the smaller the corresponding state score.

[0074] The task cost of each AGV20 can represent the time cost and energy cost required for the corresponding AGV20 to travel from its current position to the receiving point of the transport task.

[0075] The greater the task cost, the smaller the corresponding status score.

[0076] The load saturation coefficient of each AGV20 can be the ratio of the number of remaining tasks of the corresponding AGV20 to the historical daily average load. The higher the load saturation coefficient, the smaller the corresponding status score, so as to avoid other AGV20s being idle and achieve cluster load balancing.

[0077] The remaining power status of each AGV20 can be used to characterize whether the remaining power of the corresponding AGV20 is lower than the safety threshold after performing the handling task; if it is lower, the corresponding status score is lower.

[0078] The vehicle type matching degree of each AGV20 can be used to characterize the degree of matching between the capability matrix of the corresponding AGV20 and the vehicle to be transported.

[0079] The capability matrix can include lifting arm type, maximum load capacity, etc.

[0080] The higher the vehicle model matching degree, the higher the corresponding status score.

[0081] The higher the health status of each AGV20, the higher its corresponding status score.

[0082] Each AGV20 can determine its health status through self-testing. If an AGV20 is in a sub-healthy state where the sensors occasionally report errors but it can still operate normally, the status score can be 0.5.

[0083] The sum of the weights of each state parameter is 1.

[0084] The weights of each state parameter can be adaptively adjusted according to the actual situation.

[0085] In peak mode, the weight of task cost can be significantly increased, while the weight of load saturation coefficient can be reduced.

[0086] In energy-saving mode, the weights of task cost and current position can be increased.

[0087] In robust mode, when multiple AGV20s are in sub-healthy condition, the vehicle matching weight and the baseline value of health status can be implicitly increased.

[0088] When the number of AGVs is N, the N AGVs with the highest task scores can be selected as the target AGVs, where N is a positive integer not less than 1.

[0089] The main AGV200 can be selected from multiple target AGVs based on task score, communication link quality, and computing power.

[0090] As can be seen from the above technical solution, this embodiment provides an optional method for selecting target AGVs. Through this method, AGV selection can be performed by comprehensively considering multiple dimensions such as idle status, current location, task cost, load saturation coefficient, battery reserve status, vehicle type matching degree, and health status, further improving the accuracy of target AGV selection.

[0091] In some embodiments of this application, the AGV screening module may include a scoring calculation unit;

[0092] The scoring calculation unit can predict the energy consumption cost and time cost of the corresponding AGV20 to reach the handling task receiving point based on the current position of each AGV20, and calculate the energy consumption cost and time cost corresponding to the same AGV20 in a weighted manner to obtain the task cost of the corresponding AGV20; the ratio between the remaining number of tasks of each AGV20 and the historical average daily load of the AGV20 is used as the load saturation coefficient; the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree and health status corresponding to the same AGV20 are weighted to obtain the task score of the corresponding AGV20.

[0093] Specifically, the time cost can be used to characterize the estimated time for the corresponding AGV20 to perform a transport task from its current position. The degree of path congestion can be taken into account during the generation of the estimated time.

[0094] Energy consumption costs can be calculated by taking into account the current battery level of the AGV20, the energy consumption to reach the handover station, the estimated energy consumption to carry out the transport task to the destination, and the energy consumption to reach the nearest charging station after the task is completed.

[0095] The corresponding state score can be determined based on the weighted calculation of the task cost.

[0096] As can be seen from the above technical solution, this embodiment provides an optional composition method for the AGV screening module. This method can further integrate dimensions such as energy consumption cost, time cost, and the number of remaining tasks to calculate a weighted score for the task, thereby achieving load balancing of the AGV cluster, avoiding excessive workload for AGV20, and improving the service life of AGV20.

[0097] In some embodiments of this application, the scheduling center 10 may include a path generation unit.

[0098] The path generation unit can acquire a target location map, analyze the target location map, determine the distribution of feasible road segments, plan a transportation route based on the distribution of feasible road segments in the target location map, analyze the transportation route, predict road segment occupancy data, and update the distribution of feasible road segments in the target location map based on the road segment occupancy data.

[0099] The target location map can be composed of a discretized uniform grid, with each grid corresponding to a grid state, indicating whether the grid is occupied.

[0100] The grid status can include various states such as idle, reserved, soft locked, and hard locked.

[0101] Multiple consecutive grids that are in an idle state can form a feasible path;

[0102] The reservation indicates that the corresponding grid will be used as a passageway in the future;

[0103] Soft locking indicates that the corresponding grid is occupied by a faulty AGV20;

[0104] Hard locking indicates that the corresponding mesh contains static faulty objects and is impassable. The mesh state of a hard-locked mesh generally does not change.

[0105] The state of some meshes can change over time.

[0106] The path generation unit can determine the distribution of feasible road segments based on the grid status of each grid in the target location map.

[0107] The path generation unit can plan the transportation path based on the starting point and ending point of the transportation, and by taking into account the distribution of feasible road segments, using a path planning algorithm.

[0108] To avoid conflicts in transport routes that are not used for transport tasks, the route generation unit can analyze the generated transport routes, predict the road segment occupancy in future time periods, generate road segment occupancy data, and update the grid status of the grid matching the road segment occupancy data to "reserved" in order to update the feasible road segment distribution on the target location map.

[0109] The target location map can be a map of parking lots, garages, or other parking facilities.

[0110] As can be seen from the above technical solution, this embodiment provides an optional composition of the dispatch center 10. Through this method, this application can avoid conflicts in the transport paths of different transport tasks and improve transport efficiency by updating the target location map.

[0111] In some embodiments of this application, the path generation unit may include a road segment occupancy prediction component.

[0112] The road occupancy prediction component can predict the occupancy time of each road segment based on all the road segments included in the transport path; according to the basic information and the size of each road segment in the target location map, it can evaluate whether the corresponding road segment can still allow other vehicles to pass after the vehicle to be transported has passed, and obtain the evaluation result of the corresponding road segment; in the target location map, the corresponding occupancy time is marked at the matching point of the road segment whose evaluation result is not allowed.

[0113] The road segment occupancy prediction component can predict the transportation time of the transportation path based on the average transportation speed and the total length of the transportation path, and use the transportation time as the occupancy time of each communication road segment; it can also calculate the matching time range of each passage road segment based on the average transportation speed, the length and priority of each passage road segment, and use it as the occupancy time of the corresponding passage road segment.

[0114] The road occupancy prediction component can assess whether the remaining size of the corresponding road segment after accommodating the vehicle to be moved is greater than the preset maximum vehicle size, based on the size of the vehicle to be moved and the size of each road segment. If yes, the assessment result is allowed; otherwise, the assessment result is not allowed.

[0115] To avoid conflicts between newly generated transport paths and historically generated unoccupied transport paths, which could lead to conflicts between AGV20s with different transport tasks, the path generation unit can also check whether there is any overlap between the newly generated transport paths and historically generated unoccupied transport paths. If so, a progressive backtracking algorithm is used again to update the lower-priority transport paths.

[0116] Specifically, the conflict point can be traced back to the nearest feasible path branch. Different branch routes can be explored in the space following that branch. The conflict information (location, time) is recorded in a temporary conflict heatmap to prevent other tasks from being planned to this high-risk area in the short term.

[0117] The AGV20 supports automatic charging. When the battery level falls below a safe threshold, it will autonomously head to a charging station to recharge, and automatically return to the work queue after charging is complete. The vehicle handling system has a self-diagnostic function. If an AGV20 malfunctions, it will immediately report to the dispatch center, which will mark it as unavailable and reassign its unfinished tasks to other AGV20s.

[0118] As can be seen from the above technical solution, this embodiment provides an optional composition of the path generation unit. This method can further update the target location map, ensuring that the transport paths of different transport tasks do not conflict.

[0119] In some embodiments of this application, the path generation unit may further include a fault occupancy update component;

[0120] The fault occupancy update component responds to the road segment occupancy status reported by the AGV20 that has stopped operating due to a fault, and updates the target location map based on the road segment occupancy status to distribute feasible road segments.

[0121] Specifically, the fault occupancy update component can update the grid status of the grid that matches the road segment occupancy status in the target location map to soft lock, thereby updating the distribution of feasible road segments.

[0122] You can check if the transport path involves a grid that matches the road segment occupancy status. If so, you can use a progressive backtracking algorithm to update the transport path.

[0123] As can be seen from the above technical solution, this embodiment provides another optional composition of the path generation unit. This composition can address scenarios where a faulty AGV20 renders a road segment impassable, thus improving the fault tolerance of this application.

[0124] In some embodiments of this application, the dispatch center 10 may further include a moving target generation unit;

[0125] The motion target generation unit can determine a theoretical position matching the latest first position information from the transport path, select the next position of the theoretical position from the transport path as the first motion endpoint, analyze the constraint factors corresponding to the arrival of the first motion endpoint from the latest first position information, determine the first motion linear velocity and the first motion angular velocity, and generate a first motion target including the first motion endpoint, the first motion linear velocity, and the first motion angular velocity.

[0126] Analyze the obstacle constraints, velocity threshold constraints, and forward direction constraints corresponding to the arrival of the first motion endpoint from the latest first position information to determine the first motion linear velocity and the first motion angular velocity.

[0127] The forward direction is constrained by whether a turn is required. If a turn is required, the first angular velocity is not 0; if a turn is not required, the first angular velocity is 0.

[0128] As can be seen from the above technical solution, this embodiment provides an optional composition of the scheduling center 10. Through this method, the first motion target can be determined by comprehensively considering dimensions such as the transport path and physical constraints, making the phased target determination of the scheduling center 10 more reliable.

[0129] In some embodiments of this application, the main AGV200 may further include an attitude adjustment generation unit.

[0130] The attitude adjustment generation unit determines the coordinate position of each target AGV relative to the vehicle center; based on the coordinate position of each target AGV and the first linear velocity and first angular velocity of motion, it generates the forward velocity component along the vehicle's forward direction and the left and right velocity components along the vehicle's left and right directions for the corresponding target AGV; based on the forward velocity component and the left and right velocity components of each target AGV, it calculates the target running speed and target steering angle of the corresponding target AGV; based on the target running speed and target steering angle of each target AGV, it generates the first attitude adjustment state of the corresponding target AGV.

[0131] The main AGV200 for the same handling task at different times can be different AGV20s. The dispatch center 10 can select the main AGV200 from the various AGV20s for the same handling task by comprehensively considering the task score, signal strength, power consumption and computing power.

[0132] The attitude adjustment generation unit constructs a coordinate axis with the center of the vehicle to be transported as the origin, the forward direction as the positive x-axis, and the left direction perpendicular to the forward direction as the positive y-axis; the coordinates of the center point of each target AGV on the coordinate axis are used as the coordinate position of the corresponding target AGV.

[0133] The speed components of the main AGV200 and the slave AGV201 can be calculated based on the component calculation function.

[0134] The component calculation function can be shown below:

[0135]

[0136] Among them, V ix V represents the forward speed component of the i-th target AGV; iy V represents the left and right velocity components of the i-th target AGV. center ω is the linear velocity of the first motion; ω is the angular velocity of the first motion; y i x is the y-axis coordinate of the i-th target AGV. i Let x be the x-axis coordinate of the i-th target AGV.

[0137] The i-th target AGV can be either the master AGV200 or the slave AGV201.

[0138] When the vehicle to be transported turns left, ω is positive; when the vehicle to be transported turns right, ω is negative.

[0139] The square root of the sum of the squares of the forward velocity component and the left and right velocity components of the i-th target AGV can be used as the target running speed of the i-th target AGV. The target steering angle of the i-th target AGV can be calculated by the arctangent function atan2, and the first attitude adjustment state of the i-th target AGV can be generated.

[0140] The master AGV200 can transmit each first attitude adjustment state to the matched slave AGV201 through a highly reliable, low-latency wireless mesh network.

[0141] As can be seen from the above technical solution, this embodiment provides an optional composition method for the master AGV 200. By referring to the vector direction of each target AGV relative to the vehicle center, the first moving target can be decomposed into the first posture adjustment state of each target AGV, reducing the computational load of each slave AGV 201 and improving collaborative efficiency.

[0142] In some embodiments of this application, the attitude adjustment generation unit may include an attitude decomposition component.

[0143] The attitude decomposition component can determine the total attitude deviation of the corresponding target AGV by combining the target running speed, target turning angle, and current motion attitude of each target AGV. The corresponding total attitude deviation is then added to the current motion attitude of each target AGV in the same proportion to obtain the first attitude adjustment state of the corresponding target AGV, thereby updating the current motion attitude of each target AGV. The process of adding the corresponding total attitude deviation to the current motion attitude of each target AGV in the same proportion to obtain the first attitude adjustment state of the corresponding target AGV is repeated until the current motion attitude of each target AGV meets the corresponding target running speed and target turning angle.

[0144] Specifically, the attitude decomposition component can adjust the current motion attitude of AGV201 to the target running speed and target steering angle of AGV201 through multiple adjustments, thereby driving the vehicle to be transported to the first motion endpoint.

[0145] In the same adjustment, the proportion of each target AGV superimposed is the same.

[0146] As can be seen from the above technical solution, this embodiment provides an optional method for attitude adjustment generation unit. Through this method, the running speed and angular velocity of the target AGV can be gradually adjusted, avoiding sudden changes that could cause vehicle tilting and malfunction; the change ratios of different target AGVs are the same during the same superposition process, improving the coordination of changes among the various target AGVs.

[0147] When the LiDAR, vision sensor, or AGV201 reports a vehicle tilt such as roll, pitch angle abnormality, or load abnormality, the main AGV200 can activate the correction algorithm to generate fine-tuning parameters. The fine-tuning parameters mainly include the speed and steering angle of each target AGV, and sometimes also the height of the lifting mechanism.

[0148] The main AGV200 can use an IMU to measure roll angle, rotation angle, and pitch angle, compare them with the target attitude such as horizontal attitude, and generate an error signal.

[0149] Using a PID controller or a torque distribution-based method, the required parameter adjustments for each target AGV are calculated based on the error signal.

[0150] For example, if the roll angle indicates that the vehicle is tilted, the height of the lifting mechanism of the target AGV can be adjusted;

[0151] In addition, when the master AGV200 receives a load anomaly from the slave AGV201, the master AGV200 can update the first attitude adjustment state of each slave AGV201 to reduce the burden on that AGV.

[0152] For example, if AGV201 is overloaded, its speed can be reduced to alleviate its overall stress. This can be achieved by setting a gain value, calculating the ratio of the gain value to its error signal, and reducing this ratio in its speed settings to reduce overall stress. Similarly, to compensate for tilting, target AGVs in the same horizontal direction and diagonally opposite directions also need to reduce their speed, while target AGVs on the same side need to increase their speed.

[0153] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0154] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0155] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. The various embodiments of this application can be combined with each other. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A vehicle handling system, characterized in that, Including the dispatch center and multiple AGVs; The dispatch center determines the basic information of the vehicle to be transported and generates a transport task containing the transport path; it selects all target AGVs from the various AGVs to perform the transport task; when the number of target AGVs is not 1, it selects a master AGV from the target AGVs and uses the other target AGVs as slave AGVs; it receives the latest first location information uploaded by the master AGV and generates a first motion target for the vehicle to be transported based on the latest first location information and the transport path. The master AGV receives the latest first motion target issued by the scheduling center, and generates a first posture adjustment state for each slave AGV based on the latest first motion target; it issues a corresponding first posture adjustment state to each slave AGV so that each slave AGV and the master AGV can work together to drive the vehicle to be transported to achieve the first motion target; when the progress of the vehicle to be transported toward the first motion target reaches a preset progress threshold, it generates and uploads the first position information of the vehicle to be transported.

2. Vehicle handling system according to claim 1, characterized in that The scheduling center also includes an AGV screening module; The AGV screening module determines the number of AGVs based on the handling task; calculates a task score based on the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree, and health status of each AGV; and selects target AGVs that match the number of AGVs from the AGVs according to the corresponding task score.

3. Vehicle handling system according to claim 2, characterized in that The AGV screening module includes a scoring calculation unit; The scoring calculation unit predicts the energy consumption cost and time cost of the corresponding AGV to reach the handling task receiving point based on the current position of each AGV, and calculates the energy consumption cost and time cost corresponding to the same AGV in a weighted manner to obtain the task cost of the corresponding AGV; the ratio between the remaining number of tasks of each AGV and the historical average daily load of the AGV is used as the load saturation coefficient; the idle status, current position, task cost, load saturation coefficient, power retention status, vehicle type matching degree and health status corresponding to the same AGV are weighted to obtain the task score of the corresponding AGV.

4. The vehicle handling system of claim 1, wherein, The scheduling center includes a path generation unit; The path generation unit acquires a target location map and analyzes the target location map to determine the distribution of feasible road segments; Based on the distribution of feasible road segments in the target location map, a transport route is planned; the transport route is analyzed, road segment occupancy data is predicted, and the distribution of feasible road segments in the target location map is updated based on the road segment occupancy data.

5. The vehicle handling system according to claim 4, characterized in that, The path generation unit includes a road segment occupancy prediction component; The road segment occupancy prediction component predicts the occupancy time of each road segment based on all the road segments included in the transport path. Based on the basic information and the dimensions of each passage segment in the target location map, assess whether the corresponding passage segment can still allow other vehicles to pass after the vehicle to be transported has passed, and obtain the assessment result of the corresponding passage segment; mark the corresponding occupancy time at the matching point of the passage segment whose assessment result is not allowed in the target location map.

6. The vehicle handling system according to claim 4, characterized in that, The path generation unit also includes a fault occupancy update component; The fault occupancy update component responds to the road segment occupancy status reported by the AGV that has stopped operating due to a fault, and updates the target location map based on the road segment occupancy status to distribute feasible road segments.

7. The vehicle handling system according to claim 1, characterized in that, The scheduling center also includes a moving target generation unit; The motion target generation unit determines a theoretical position matching the latest first position information from the transport path, selects the next position of the theoretical position from the transport path as the first motion endpoint, analyzes the constraint factors corresponding to the arrival of the first motion endpoint from the latest first position information, determines the first motion linear velocity and the first motion angular velocity, and generates a first motion target including the first motion endpoint, the first motion linear velocity, and the first motion angular velocity.

8. The vehicle handling system according to claim 7, characterized in that, The main AGV also includes an attitude adjustment generation unit; The attitude adjustment generation unit determines the coordinate position of each target AGV relative to the vehicle center; based on the coordinate position of each target AGV and the first linear velocity and first angular velocity of motion, it generates the forward velocity component along the vehicle's forward direction and the left and right velocity components along the vehicle's left and right directions for the corresponding target AGV; based on the forward velocity component and the left and right velocity components of each target AGV, it calculates the target running speed and target steering angle of the corresponding target AGV; based on the target running speed and target steering angle of each target AGV, it generates the first attitude adjustment state of the corresponding target AGV.

9. The vehicle handling system according to claim 8, characterized in that, The attitude adjustment generation unit includes an attitude decomposition component; The attitude decomposition component determines the total attitude deviation of each target AGV by combining the target running speed, target turning angle, and current motion attitude of each target AGV. It then adds the corresponding total attitude deviation to the current motion attitude of each target AGV in the same proportion to obtain the first attitude adjustment state of the corresponding target AGV, thereby updating the current motion attitude of each target AGV. The process continues until the current motion attitude of each target AGV satisfies the corresponding target running speed and target turning angle.

Images