Multi-intelligent carrying AGV path planning method, device and medium

Abstract

This application discloses a method, device, and medium for path planning of multiple intelligent transport AGVs, relating to the field of industrial control technology. The method includes: responding to a path planning request, using the target AGV's starting path point as the first current path point based on map topology data; searching for adjacent reachable path points, and determining a first actual travel cost based on the spatiotemporal overlap information of the sub-paths corresponding to each adjacent reachable path point with other AGV planned paths; adding the first cumulative cost of the first current path point to the first actual travel cost to obtain a first cumulative cost, and combining this with a first heuristic cost to obtain a first evaluated cost; storing adjacent reachable path points in an open list and removing the current path point, selecting the one with the smallest evaluated cost as the new first current path point, and iterating in this way until the termination path point is reached, backtracking to generate a globally planned path and adding a locking flag. Thus, iterative path search can effectively improve road utilization and the overall operating efficiency of multi-AGV systems.

Description

Technical Field

[0001] This application relates to the field of industrial control technology, and in particular to a method, equipment and medium for path planning of multi-intelligent handling AGVs. Background Technology

[0002] With the large-scale development of discrete manufacturing and intelligent warehousing industries, multi-AGV clusters have become the execution carriers for material transfer, process flow, and warehousing sorting in workshops. The quality of trajectory planning schemes will directly determine the overall throughput capacity of workshop logistics, the degree of equipment wear and tear, and the stability of continuous operation of the entire production line.

[0003] In related technologies, the mature AGV path planning system in the industry is mainly divided into three technical approaches: independent trajectory solving for a single vehicle, centralized time window collaborative scheduling, and local autonomous obstacle avoidance for a single machine. However, when faced with complex working conditions such as large-scale clusters of dozens of vehicles and real-time changes in on-site obstacles, these three technical solutions cannot simultaneously meet the core requirements of multi-device collaborative management, real-time dynamic response, global path optimization, and low-computing-power lightweight operation. Summary of the Invention

[0004] This application provides a method, device, and medium for planning the path of multiple intelligent handling AGVs, in order to solve the following technical problem: how to effectively improve road utilization and the overall operating efficiency of a multi-AGV system.

[0005] In a first aspect, embodiments of this application provide a multi-intelligent handling AGV path planning method, the method comprising: In response to receiving a path planning request from a target AGV, based on pre-built map topology data, the starting path point of the target AGV is taken as the first current path point, and a path search is performed on the first current path point until the ending path point is reached: Find all first adjacent reachable path points of the first current path point; For each of the first adjacent reachable path points, the road segment between the first adjacent reachable path point and the first current path point is taken as the first sub-path to be evaluated corresponding to the first adjacent reachable path point. Based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first actual driving cost of the first adjacent reachable path point is determined. For each of the first adjacent reachable path points, the first cumulative cost of the first current path point is added to the first actual driving cost of the first adjacent reachable path point to obtain the first cumulative cost of the first adjacent reachable path point, and the first cumulative cost of the first adjacent reachable path point is added to the first heuristic cost of the first adjacent reachable path point to obtain the first evaluation cost of the first adjacent reachable path point. Store each of the first adjacent reachable path points in the first open list, and delete the first current path point from the first open list; The path point corresponding to the minimum first evaluation cost in the first open list is taken as the new first current path point; In response to the new first current path point being the termination path point, a global planned path from the starting path point to the termination path point is backtracked to enable the target AGV to travel along the global planned path, and a lock flag is added to the global planned path.

[0006] Secondly, embodiments of this application also provide a multi-intelligent transport AGV path planning device, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a multi-intelligent transport AGV path planning method as described above.

[0007] Thirdly, embodiments of this application also provide a computer storage medium storing computer-executable instructions, which, when executed, implement a multi-intelligent handling AGV path planning method as described above.

[0008] The multi-intelligent handling AGV path planning method, equipment, and medium provided in this application have the following beneficial effects: First, upon receiving the path planning request from the target AGV, a search is initiated using the starting path point as the first current path point, querying all first adjacent reachable path points to ensure the completeness of the search space. Next, based on the spatiotemporal overlap information between the first sub-path to be evaluated and other AGV planned paths, the actual travel cost of the first adjacent reachable path point is determined. This expands path evaluation from simple geometric distance to the spatiotemporal dimension of multi-vehicle collaboration, allowing proactive avoidance of high-conflict-risk road sections during the planning phase to balance traffic flow across various channels in the factory area. Subsequently, the first cumulative cost of the first current path point is added to the actual travel cost to obtain the first cumulative cost of the first adjacent reachable path point. This, combined with the first heuristic cost, generates the first evaluated cost of the first reachable path point. This ensures that the cost from the starting point to the current node can be correctly passed layer by layer along the search path, and that the heuristic term guides the search direction towards the target point, avoiding blind expansion. Finally, the calculated cost of the first... By storing adjacent reachable path points in the first open list and removing processed current nodes, dynamic maintenance of the search front and effective management of traversed nodes can be achieved, preventing redundant calculations and path loops. By selecting the node with the lowest evaluation cost from the open list as the new first current path point, each iteration prioritizes the direction with the lowest overall cost, ensuring that the final output global planning path approaches the global optimum while considering multi-vehicle conflict penalties. When the new current path reaches the termination path point, the global planning path is backtracked to generate a global planning path, transforming the optimal node sequence accumulated throughout the iteration process into a specific driving route that the target AGV can execute. A locking flag is added to the global planning path to provide real-time structured storage of road segment resource occupancy status, providing a direct data foundation for subsequent conflict-aware path planning for other AGVs and vehicle locking detection and dynamic replanning during the operation phase. This can effectively improve the efficiency of multi-AGV cluster collaborative operations and road utilization. Attached Figure Description

[0009] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart of a multi-intelligent handling AGV path planning method provided in this application embodiment; Figure 2 This is a schematic diagram of the internal structure of a multi-intelligent handling AGV path planning device provided in an embodiment of this application. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] It is understood that in the embodiments of this application, data related to user information (such as user accounts) is involved. When the embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with relevant laws, regulations and standards.

[0012] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0013] In the following description, the terms “first, second, ...” are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that “first, second, ...” may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0014] With the large-scale development of discrete manufacturing and intelligent warehousing industries, multi-AGV clusters have become the execution carriers for material transfer, process flow, and warehousing sorting in workshops. The quality of trajectory planning schemes will directly determine the overall throughput capacity of workshop logistics, the degree of equipment wear and tear, and the stability of continuous operation of the entire production line.

[0015] In related technologies, the mature AGV path planning systems in the industry are mainly divided into three technical approaches: independent trajectory solving for a single vehicle, centralized time window collaborative scheduling, and local autonomous obstacle avoidance for a single machine. However, when faced with large-scale clusters of dozens of vehicles and complex working conditions with real-time changes in on-site obstacles, these three technical solutions all have the following inherent technical defects that are difficult to eradicate: (1) Single-device independent global search algorithms have a limitation in adaptability. In related technologies, static graph search algorithms (such as Dijkstra's algorithm, standard A) have limitations in adaptability. Basic swarm intelligence optimization algorithms (such as ant colony optimization, particle swarm optimization, etc.) only take the geometric travel distance of a single AGV as the only optimization index in the modeling and solving stage. The modeling process assumes that the working environment is fixed and unchanged, and there is no interaction interference from other transfer equipment. The coupling constraints of multi-vehicle synchronous operation can be completely ignored. (2) In the case of a large-scale cluster of complex production lines with dense intersections and uncertain workstation parking time, the centralized architecture will explode with the expansion of equipment and the amount of computing power. The static time window and the time sequence of workstation operation are mismatched, which can easily lead to long-term road segment path locking, traffic congestion, and deadlock between directions. Furthermore, there is no automatic unlocking and replanning method after a fault conflict, and there is also the risk of single point failure of the server. (3) In the complex production line large-scale cluster scenario with dense intersections and uncertain workstation parking time, the single-machine obstacle avoidance of the multi-AGV management scheme lacks global traffic flow collaborative management and can only temporarily avoid obstacles locally. It cannot predict global path occupation conflicts. When the same direction or the path resource is locked, it cannot release the road segment autonomously and re-plan the optimal path, resulting in a significant decrease in the overall transfer efficiency of the cluster.

[0016] In summary, existing AGV path planning technologies cannot simultaneously meet the four core requirements of multi-device collaborative management, real-time dynamic response, global path optimization, and low-computing-power lightweight operation.

[0017] Based on this, the embodiments of this application provide a multi-intelligent handling AGV path planning method, which can effectively improve road utilization and the overall operating efficiency of multi-AGV systems.

[0018] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0019] Figure 1 This application provides a flowchart of a multi-AGV path planning method. This method can be applied to various multi-AGV path planning scenarios, such as discrete manufacturing workshops, intelligent warehousing and sorting centers, logistics parks / container terminals, and automated factories. Certain input parameters or intermediate results in the process can be manually adjusted to help improve accuracy.

[0020] This application provides a multi-intelligent material handling AGV path planning method. It should be noted that the execution entity in this embodiment can be a server or any terminal device with data processing capabilities. For example, the server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. The terminal device can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, vehicle terminal, etc., but is not limited to these.

[0021] like Figure 1 As shown in the figure, the multi-intelligent handling AGV path planning method provided in this application embodiment specifically includes the following steps: Step 101: In response to receiving the path planning request from the target AGV, the starting path point of the target AGV is taken as the first current path point based on the pre-built map topology data.

[0022] Here, the pre-built map topology data refers to a set of structured road network data obtained by abstracting and modeling the actual environment of the factory area before the AGV system goes online and runs. Preferably, the pre-built map topology data can adopt a point-line topology structure, with each path point storing coordinates, basic road segment traffic cost, road width, slope attributes, etc.

[0023] Step 102: Perform a path search on the first current path point until the termination path point is reached.

[0024] Here, the execution of step 102 includes iteratively executing steps 1021 to 1026.

[0025] It should be noted that path search can employ improved A The pathfinding algorithm pre-initializes the first open list before path searching, storing the starting path point in the first open list; the improved A Pathfinding algorithms are based on classic A Based on the pathfinding algorithm, a path search algorithm is obtained by incorporating the real-time planning path information of other AGVs in the field into the cost evaluation process of node expansion.

[0026] Step 1021: Find all first adjacent reachable path points of the first current path point.

[0027] Here, the first adjacent reachable path point refers to a path point that can be directly reached from the first current path point along a road segment without any physical obstacles or traffic restrictions.

[0028] Step 1022: For each of the first adjacent reachable path points, the road segment between the first adjacent reachable path point and the first current path point is taken as the first sub-path to be evaluated corresponding to the first adjacent reachable path point. Based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first actual driving cost of the first adjacent reachable path point is determined.

[0029] In some embodiments, the determination of the first actual travel cost of the first adjacent reachable path point based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario in step 1022 above can be achieved in the following way: based on the first basic cost mapping rule, the passage attributes of the first sub-path to be evaluated corresponding to the first adjacent reachable path point are mapped to obtain the first basic travel cost of the first adjacent reachable path point, wherein the passage attributes include at least one of road segment length, road segment width, and road surface condition; in response to the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the determination of the first actual travel cost of the first adjacent reachable path point ... and the planned path information of other AGVs in the preset scenario can be mapped to obtain the first basic travel cost of the first adjacent reachable path point. Since there are no overlapping segments between the planned paths of the GVs, the first basic driving cost of the first adjacent reachable path point is taken as the first actual driving cost of the first adjacent reachable path point. In response to the existence of overlapping segments between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned paths of other AGVs in the preset scenario, the first conflict penalty cost of the first adjacent reachable path point is determined based on the driving direction of the target AGV and other AGVs in the preset scenario in the overlapping segment. The first basic driving cost of the first adjacent reachable path point is added to the first conflict penalty cost of the first adjacent reachable path point to obtain the first actual driving cost of the first adjacent reachable path point.

[0030] Thus, by comparing the spatiotemporal overlap between the sub-segment to be evaluated and other planned paths of AGVs in real time at each step of the path search, the passage cost is dynamically decomposed into two parts: static basic cost and dynamic conflict penalty. This upgrades path evaluation from a single geometric distance dimension to a spatiotemporal game dimension of multi-vehicle collaboration. When there is no overlap, the basic cost is directly used to ensure lightweight computation of the algorithm on idle road segments. When there is overlap, differentiated penalties are applied to the same-direction and opposite-direction travel directions, allowing the algorithm to quantify the risk level of different conflict types during the planning stage. This proactively guides the target AGV to prioritize same-direction traffic segments and avoid high-risk areas of opposite-direction intersections, reducing the probability of locking and blocking in subsequent operations from the source. Through this mechanism of embedding global traffic flow prediction into the node expansion process, without relying on centralized time window scheduling or increasing the computing power burden of a single vehicle, a balanced allocation of passage resources for multi-AGV clusters and optimal approximation of global path quality can be achieved. At the same time, the utilization rate of factory roads and the overall operating efficiency of the system can be improved.

[0031] Here, the first basic cost mapping rule is obtained by pre-calibrating the path length, road width, and road surface conditions. For example, it can be a linear function or other mapping functions, without specific limitations.

[0032] As an example, taking the actual operation of an electronic assembly workshop as an example, there is a 10-meter-long straight passage between path point A and path point B on the factory map. The first basic travel cost of this section is 10. The target AGV travels from A to B. At this time, there is another working AGV's planned path, which also happens to include the section from A to B. That is, the sub-path to be evaluated has a spatiotemporal overlap with other AGVs. Further comparison shows that the path point sequence order of the two AGVs on this section is from A to B, and the travel direction is the same. That is, the current conflict is a same-direction conflict. The first preset penalty cost (for example, 5) is added to the first basic travel cost of 10, and the first actual travel cost of path point B is 15.

[0033] In another example, taking the actual operation of an electronic assembly workshop as an example, there is a 15-meter-long straight passage between path point C and path point D in the factory map. The first basic travel cost of this section is 15. There is no spatiotemporal overlap between the planned paths of other AGVs and the target AGV from C to D. Therefore, the first travel cost of path D is equal to the first basic travel cost of 15.

[0034] In some embodiments, the determination of the first conflict penalty cost of the first adjacent reachable path point based on the travel directions of the target AGV and other AGVs in the preset scene within the overlapping segment can be achieved in the following manner: Counting the first number of times the target AGV and other AGVs in the preset scene travel in the same direction within the overlapping segment, and multiplying the first number by the first preset penalty cost to obtain the first same-direction conflict penalty sub-cost of the first adjacent reachable path point; Counting the second number of times the target AGV and other AGVs in the preset scene travel in opposite directions within the overlapping segment, and multiplying the second number by the second preset penalty cost to obtain the first opposite-direction conflict penalty sub-cost of the first adjacent reachable path point; Adding the first same-direction conflict penalty sub-cost of the first adjacent reachable path point to the first opposite-direction conflict penalty sub-cost of the first adjacent reachable path point to obtain the first conflict penalty cost of the first adjacent reachable path point.

[0035] Thus, by separately counting the number of other AGVs traveling in the same direction and opposite direction as the target AGV within the overlapping segment, and then multiplying each by its corresponding preset penalty cost and summing them up, refined cost quantification can be achieved in multi-vehicle conflict scenarios. This quantification design allows the conflict penalty to increase linearly with the actual number of concurrent conflicting vehicles, thereby accurately reflecting the true traffic risk level when a road segment is shared by multiple vehicles. Differentiated unit prices are adopted for same-direction and opposite-direction traffic, which not only reflects the objective physical law that the risk of collisions due to opposite-direction collisions is significantly higher than that of deceleration due to following vehicles in the same direction, but also guides the path search algorithm to actively avoid high-risk channels where multiple vehicles converge in opposite directions in complex and dense traffic flow through the dual dimensions of quantity and unit price, while allowing the selection of road segments with dense same-direction traffic flow but lower risk when necessary. The cost calculation mechanism of this scheme transforms the coupling constraints of multiple AGVs into an accumulative quantified cost signal without increasing the algorithm complexity, enabling the path search results to distribute traffic resources more evenly at the global level, reducing the probability of vehicle locking and blocking from the source, thereby effectively improving the overall transfer efficiency of the cluster.

[0036] Here, since the same-direction conflict only has the risk of deceleration of following vehicles, while the reverse conflict has the risk of head-on collision and complete lock-up, the reverse conflict needs to be set with a higher penalty cost. The penalty cost of the first same-direction conflict and the penalty cost of the first reverse conflict can be determined according to different specific scenarios. Preferably, the penalty cost of the first same-direction conflict is one time the preset conflict penalty cost, and the penalty cost of the first reverse conflict is three times the preset conflict penalty cost.

[0037] As an example, consider a two-way main passage in a smart warehouse. This passage connects path point E and path point F, with a first basic travel cost of 8. Then, the system detects that this section of the road overlaps with the planned paths of three other AGVs in both directions. The first AGV also travels from E to F in the same direction; the second AGV travels from F to E in the opposite direction; and the third AGV also travels from F to E in the opposite direction. Subsequently, the number of vehicles traveling in the same direction is counted as 1, and the number of vehicles traveling in the opposite direction is counted as 2. The first preset penalty cost is set to 3, and the second preset penalty cost is set to 10. Then, the sub-cost of the conflict in the same direction is 3, and the sub-cost of the conflict in the opposite direction is 20. The sum of the two is the first conflict penalty cost of the current section, which is 23.

[0038] Step 1023: For each of the first adjacent reachable path points, add the first cumulative cost of the first current path point to the first actual driving cost of the first adjacent reachable path point to obtain the first cumulative cost of the first adjacent reachable path point, and add the first cumulative cost of the first adjacent reachable path point to the first heuristic cost of the first adjacent reachable path point to obtain the first evaluation cost of the first adjacent reachable path point.

[0039] Here, the first cumulative cost is the sum of all basic costs and conflict penalty costs that have been traversed from the starting path point to the current path point.

[0040] As an example, the first estimated cost of the first adjacent reachable path point can be calculated according to formulas (1) and (2). .

[0041] (1) (2) in, The first cumulative cost of the first adjacent reachable path point; The first cumulative cost of the first current path point; The first actual travel cost is the first adjacent reachable path point; The first heuristic cost for the first adjacent reachable path point is the cost from the first adjacent reachable path point to the terminal path point, which can be calculated using Manhattan distance or Euclidean distance.

[0042] Step 1024: Store each of the first adjacent reachable path points in the first open list, and delete the first current path point from the first open list.

[0043] Step 1025: Take the path point with the minimum first evaluation cost in the first open list as the new first current path point.

[0044] Here, after taking the path point corresponding to the minimum first evaluation cost in the first open list as the new first current path point, if the new first current path point is not the termination path point, then the above steps 1021 to 1025 are executed with the new first current path point as the target object.

[0045] Step 1026: In response to the new first current path point being the termination path point, backtrack to obtain the global planned path from the starting path point to the termination path point, so that the target AGV travels along the global planned path, and add a lock flag to the global planned path.

[0046] Here, if the new first current path point is the termination path point, the iteration loop is terminated. The global planned path of the target AGV from the starting path point to the target path point is backtracked and output. The path locking information is written to the corresponding road segment in the map topology data for the road segment sequence and the expected travel time included in the global planned path.

[0047] It should be noted that the system maintains a locking record table for each road segment. After the global planned path is output, for each road segment between path points involved in the path, the system updates its occupancy status to locked, and simultaneously writes the occupant AGV identifier (i.e., the target AGV ID), the occupancy direction (forward or reverse), and the estimated occupancy time period (estimated based on the AGV's current speed, road segment length, and task priority). This locking identifier serves as a status marker for road segment access resources, and is used to inform other AGVs in the system about improvement measures. The algorithm provides real-time spatiotemporal path occupancy data, enabling other AGVs to query conflict information on the road segment and apply corresponding penalties when planning their own paths. On the other hand, it is used for conflict monitoring during operation. When other AGVs actually approach the road segment or resource contention occurs, the system can quickly determine whether a vehicle locking blockage state has been triggered based on the locking flag, and then perform avoidance, unlocking, and dynamic replanning operations. When the target AGV actually leaves the road segment or completes its work and moves to the designated docking point, the system automatically clears the locking record of the corresponding road segment, restores the road right-of-way, and allows other AGVs in the yard to plan their routes normally. Thus, through this locking flag mechanism—adding it immediately after path planning and automatically clearing it after use—the system can achieve refined collaborative management of multi-AGV access resources, preventing disorderly contention for road segment resources and global deadlocks from the source.

[0048] In some embodiments, after performing step 1026 above, the following processing may also be performed: real-time acquisition of the driving status of the target AGV; in response to the target AGV's driving status being locked and blocked, controlling the target AGV to drive to the nearest temporary avoidance point and releasing the lock flag of the target AGV's global planned path; based on pre-built map topology data, taking the nearest temporary avoidance point of the target AGV as the second current path point, and performing path search on the second current path point until reaching the termination path point; finding all second adjacent reachable path points of the second current path point; for each second adjacent reachable path point, taking the road segment between the second adjacent reachable path point and the second current path point as the second sub-path to be evaluated corresponding to the second adjacent reachable path point, and determining the second adjacent reachable path based on the spatiotemporal overlap information between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario. The second actual travel cost of the second adjacent reachable path point; for each second adjacent reachable path point, the second cumulative cost of the second current path point is added to the second actual travel cost of the second adjacent reachable path point to obtain the second cumulative cost of the second adjacent reachable path point, and the second cumulative cost of the second adjacent reachable path point is added to the second heuristic cost of the second adjacent reachable path point to obtain the second evaluated cost of the second adjacent reachable path point; each second adjacent reachable path point is stored in the second open list, and the second current path point is deleted from the second open list; the path point corresponding to the minimum second evaluated cost in the second open list is taken as the new second current path point; in response to the new second current path point being the termination path point, a valid planned path from the nearest temporary avoidance point to the termination path point is backtracked to enable the target AGV to continue traveling along the valid planned path, and a lock flag is added to the valid planned path.

[0049] Thus, on the one hand, the pre-set avoidance points and temporary stopping mechanisms allow opposing or same-direction traffic to quickly resume flow, avoiding production line stagnation caused by prolonged multi-vehicle standoffs in traditional deadlock scenarios; on the other hand, by removing the original path lock markers, the right-of-way of occupied road segments can be returned in a timely manner, eliminating the resource root cause of continuous vehicle locking and preventing the deadlock state from spreading to the entire road network; at the same time, the replanning process reuses and improves A The algorithm continues the conflict penalty rules that differentiate between same-direction and reverse-direction, so that the detour path can avoid the original conflict section while still taking into account the overall traffic efficiency. Even if multiple candidate paths have continuous conflicts, the algorithm can naturally select the optimal detour channel through cost ranking. This scheme seamlessly connects the real-time blocking response in the operation phase with the collaborative cost evaluation in the planning phase, and achieves adaptive recovery and dynamic load balancing of multi-AGV systems without relying on manual intervention. It can greatly improve the continuity and robustness of cluster operations under complex working conditions.

[0050] Here, after obtaining the global planned path for the target AGV through path planning, the location, driving status, path locking information, and workstation parking status of all AGVs in the field are collected in real time. The system continuously monitors whether the target AGV triggers a locking / blocking state. Data collection polls the entire AGV's operation data at fixed intervals (e.g., 500ms), simultaneously recording the occupancy / locking flags of each road segment, workstation, and avoidance point. If the target AGV is locked / blocked due to a conflicting road segment being occupied or other AGVs occupying the passageway for an extended period, a path unlocking process is executed to release the path locking flag of the currently occupied segment and clear the cache of the currently incomplete driving path. Afterwards, the system retrieves pre-built map topology data from intersections, narrow passages, and workstations. The system uses a set of dedicated avoidance points at the entrance and exit of the AGV to match the nearest temporary avoidance point based on the target AGV's current congestion location, vehicle size, and road width. The system then controls the target AGV to travel to the nearest matched temporary avoidance point, actively avoiding oncoming traffic and simultaneously releasing the two-way path lock information of the conflict section. The temporary avoidance point is a pre-planned idle parking area within the factory area, used only for temporary avoidance and will not occupy the main road, causing a global road lock. After the avoidance is completed, the nearest temporary avoidance point is used as the current path point, and the path search method from steps 1021 to 1026 is used for path replanning to obtain an effective planned path from the nearest temporary avoidance point to the termination path point, allowing the target AGV to continue traveling along the effective planned path towards the termination path point.

[0051] It should be noted that if the target AGV encounters a blockage or conflict during its journey, the above avoidance and replanning operations will be repeated until the target AGV reaches the termination path point.

[0052] In some embodiments, the determination of the second actual travel cost of the second adjacent reachable path point based on the spatiotemporal overlap information between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario can be achieved in the following way: based on the second basic cost mapping rule, the passage attributes of the second sub-path to be evaluated corresponding to the second adjacent reachable path point are mapped to obtain the second basic travel cost of the second adjacent reachable path point, wherein the passage attributes include at least one of road segment length, road segment width, and road surface condition; in response to the spatiotemporal overlap information between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the determination of the second actual travel cost of the second adjacent reachable path point ... If there are no overlapping segments between the planned path information, the second basic driving cost of the second adjacent reachable path point is used as the second actual driving cost of the second adjacent reachable path point. In response to the existence of overlapping segments between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the second conflict penalty cost of the second adjacent reachable path point is determined based on the driving direction of the target AGV and other AGVs in the preset scenario in the overlapping segment. The second basic driving cost of the second adjacent reachable path point is added to the second conflict penalty cost of the second adjacent reachable path point to obtain the second actual driving cost of the second adjacent reachable path point.

[0053] It should be noted that the calculation process for the second real driving cost is similar to that for the first real driving cost. For details, please refer to the calculation process for the first real driving cost. It will not be repeated here.

[0054] In some embodiments, the determination of the second conflict penalty cost of the second adjacent reachable path point based on the travel directions of the target AGV and other AGVs in the preset scenario within the overlapping segment can be achieved in the following manner: Counting a third number of instances where the target AGV and other AGVs in the preset scenario travel in the same direction within the overlapping segment, and multiplying the third number by a third preset penalty cost to obtain the second same-direction conflict penalty sub-cost of the second adjacent reachable path point; counting a fourth number of instances where the target AGV and other AGVs in the preset scenario travel in opposite directions within the overlapping segment, and multiplying the fourth number by a fourth preset penalty cost to obtain the second opposite-direction conflict penalty sub-cost of the second adjacent reachable path point; and adding the second same-direction conflict penalty sub-cost of the second adjacent reachable path point to the second opposite-direction conflict penalty sub-cost of the second adjacent reachable path point to obtain the second conflict penalty cost of the second adjacent reachable path point.

[0055] It should be noted that the calculation process for the second conflict penalty cost is similar to that for the first conflict penalty cost. For details, please refer to the calculation process for the first conflict penalty cost. It will not be repeated here.

[0056] In some embodiments, the following processing may also be performed: In response to the presence of a parked AGV within the preset area, based on pre-built map topology data, the parking point of the parked AGV is taken as the third current path point, and a path search is performed on the third current path point until a waiting path point is reached: all third adjacent reachable path points of the third current path point are found; for each third adjacent reachable path point, the road segment between the third adjacent reachable path point and the third current path point is taken as the third sub-path to be evaluated corresponding to the third adjacent reachable path point, and based on the spatiotemporal overlap information between the third sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the third actual travel cost of the third adjacent reachable path point is determined; for each third adjacent reachable path point, the parking point of the parked AGV is taken as the third current path point, and a path search is performed on the third current path point until a waiting path point is reached; the third adjacent reachable path point is taken as the third current path point, and the road segment between the third adjacent reachable path point and the third current path point is taken as the third sub-path to be evaluated corresponding to the third current path point, and the third actual travel cost of the third adjacent reachable path point is determined based on the spatiotemporal overlap information between the third sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario; for each third adjacent reachable path point, the third actual travel cost of the parked AGV is taken as the third current path point. The third cumulative cost of the third current path point is added to the third actual driving cost of the third adjacent reachable path point to obtain the third cumulative cost of the third adjacent reachable path point. The third cumulative cost of the third adjacent reachable path point is then added to the third heuristic cost of the third adjacent reachable path point to obtain the third evaluated cost of the third adjacent reachable path point. Each of the third adjacent reachable path points is stored in a third open list, and the third current path point is deleted from the third open list. The path point with the minimum third evaluated cost in the third open list is taken as the new third current path point. In response to the new third current path point being the waiting path point, a transfer planning path from the starting path point to the waiting path point is backtracked to enable the parked AGV to transfer to the waiting path point along the transfer planning path.

[0057] In this way, by transforming passive waiting into proactive scheduling, it is possible to avoid chain congestion caused by parked vehicles occupying key passages or workstation entrances and exits for extended periods. The planning of transfer routes also incorporates the assessment of the cost of spatiotemporal conflicts, ensuring that the removal process itself will not create new deadlocks with AGVs operating normally within the site. After the parked AGV is transferred to the preset waiting point, the locking markers on the original occupied road sections and workstations are cleared, and the right-of-way is restored immediately, allowing subsequent vehicles to pass through normally. The conflict handling mechanism in this solution can work in conjunction with conflict unlocking and replanning to solve resource occupation problems in two different scenarios: dynamic driving congestion and static parking congestion. This will jointly ensure the overall smooth flow of factory roads and the continuous and efficient operation of the multi-AGV system.

[0058] Here, if there is a parked AGV in the preset work area, the parking point of the parked AGV is used as the current path point. The path search method in steps 1021 to 1026 is used to search for the path and obtain the transfer planning path from the parking point to the preset waiting path point. This allows the parked AGV to transfer to the preset waiting path point along the transfer planning path, thereby releasing the passage resources of the main road and work point and eliminating the root cause of continuous vehicle locking. The preset waiting path point can be a preset idle parking area or a backup waiting point.

[0059] In some embodiments, the determination of the third actual travel cost of the third adjacent reachable path point based on the spatiotemporal overlap information between the third sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario can be achieved in the following way: based on the third basic cost mapping rule, the passage attributes of the third sub-path to be evaluated corresponding to the third adjacent reachable path point are mapped to obtain the third basic travel cost of the third adjacent reachable path point, wherein the passage attributes include at least one of road segment length, road segment width, and road surface condition; in response to the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the third actual travel cost of the third adjacent reachable path point is determined. If there are no overlapping segments between the path information, the third basic driving cost of the third adjacent reachable path point is used as the third actual driving cost of the third adjacent reachable path point. In response to the existence of overlapping segments between the third sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the third conflict penalty cost of the third adjacent reachable path point is determined based on the driving direction of the target AGV and other AGVs in the preset scenario in the overlapping segment. The third basic driving cost of the third adjacent reachable path point is added to the third conflict penalty cost of the first adjacent reachable path point to obtain the third actual driving cost of the third adjacent reachable path point.

[0060] It should be noted that the calculation process for the third real driving cost is similar to that for the first real driving cost. For details, please refer to the calculation process for the first real driving cost. It will not be repeated here.

[0061] In some embodiments, the determination of the third conflict penalty cost of the third adjacent reachable path point based on the travel directions of the target AGV and other AGVs in the preset scenario within the overlapping segment can be achieved in the following manner: Count the fifth number of times the target AGV and other AGVs in the preset scenario travel in the same direction within the overlapping segment, and multiply the fifth number by the fifth preset penalty cost to obtain the third same-direction conflict penalty sub-cost of the third adjacent reachable path point; count the sixth number of times the target AGV and other AGVs in the preset scenario travel in opposite directions within the overlapping segment, and multiply the sixth number by the sixth preset penalty cost to obtain the third opposite-direction conflict penalty sub-cost of the third adjacent reachable path point; add the third same-direction conflict penalty sub-cost of the third adjacent reachable path point to the third opposite-direction conflict penalty sub-cost of the third adjacent reachable path point to obtain the third conflict penalty cost of the third adjacent reachable path point.

[0062] It should be noted that the calculation process for the third conflict penalty cost is similar to that for the first conflict penalty cost. For details, please refer to the calculation process for the first conflict penalty cost. It will not be repeated here.

[0063] In some embodiments, the following processing may also be performed: in response to other AGVs that cause a locking conflict leaving the locked area or completing their workstation operations and moving to a designated docking point, the locking record of the corresponding road segment is automatically cleared, the right-of-way of the road segment is restored, and the remaining AGVs in the yard are allowed to pass normally.

[0064] This avoids false road occupancy caused by long-term residual locking markers, preventing the system from continuously misjudging the road segment as impassable and incurring unnecessary detour costs; it also allows for improvements to other AGVs. The algorithm can promptly detect the restoration of right-of-way and reinstate the released road segments into the candidate range of the optimal path, thereby improving the quality of global path search and the utilization rate of road resources. This mechanism complements active unlocking and avoidance replanning, covering both the natural recovery scenario of vehicles autonomously leaving and ensuring that the road network status remains synchronized with the actual situation on site, thus guaranteeing the self-consistency and long-term operational stability of the multi-AGV collaborative scheduling system from the bottom layer.

[0065] Here, when other AGVs that experienced locking conflicts leave the locked area, complete their work at their workstations, and move to their designated parking points, the system will automatically clear the locking record for the corresponding road segment, restore the right-of-way, and allow other AGVs in the area to pass normally. The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide a multi-intelligent handling AGV path planning device, the structure of which is as follows: Figure 2 As shown.

[0066] Figure 2 This is a schematic diagram of the internal structure of a multi-intelligent handling AGV path planning device provided in an embodiment of this application. Figure 2 As shown, the device includes: At least one processor 201; And a memory 202 that is communicatively connected to at least one processor; The memory 202 stores instructions executable by at least one processor, which are executed by at least one processor 201 to enable at least one processor 201 to: In response to receiving a path planning request from a target AGV, based on pre-built map topology data, the starting path point of the target AGV is taken as the first current path point, and a path search is performed on the first current path point until the ending path point is reached: Find all first adjacent reachable path points of the first current path point; For each of the first adjacent reachable path points, the road segment between the first adjacent reachable path point and the first current path point is taken as the first sub-path to be evaluated corresponding to the first adjacent reachable path point. Based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first actual driving cost of the first adjacent reachable path point is determined. For each of the first adjacent reachable path points, the first cumulative cost of the first current path point is added to the first actual driving cost of the first adjacent reachable path point to obtain the first cumulative cost of the first adjacent reachable path point, and the first cumulative cost of the first adjacent reachable path point is added to the first heuristic cost of the first adjacent reachable path point to obtain the first evaluation cost of the first adjacent reachable path point. Store each of the first adjacent reachable path points in the first open list, and delete the first current path point from the first open list; The path point corresponding to the minimum first evaluation cost in the first open list is taken as the new first current path point; In response to the new first current path point being the termination path point, a global planned path from the starting path point to the termination path point is backtracked to enable the target AGV to travel along the global planned path, and a lock flag is added to the global planned path.

[0067] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured as follows: In response to receiving a path planning request from a target AGV, based on pre-built map topology data, the starting path point of the target AGV is taken as the first current path point, and a path search is performed on the first current path point until the ending path point is reached: Find all first adjacent reachable path points of the first current path point; For each of the first adjacent reachable path points, the road segment between the first adjacent reachable path point and the first current path point is taken as the first sub-path to be evaluated corresponding to the first adjacent reachable path point. Based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first actual driving cost of the first adjacent reachable path point is determined. For each of the first adjacent reachable path points, the first cumulative cost of the first current path point is added to the first actual driving cost of the first adjacent reachable path point to obtain the first cumulative cost of the first adjacent reachable path point, and the first cumulative cost of the first adjacent reachable path point is added to the first heuristic cost of the first adjacent reachable path point to obtain the first evaluation cost of the first adjacent reachable path point. Store each of the first adjacent reachable path points in the first open list, and delete the first current path point from the first open list; The path point corresponding to the minimum first evaluation cost in the first open list is taken as the new first current path point; In response to the new first current path point being the termination path point, a global planned path from the starting path point to the termination path point is backtracked to enable the target AGV to travel along the global planned path, and a lock flag is added to the global planned path.

[0068] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0069] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0070] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0072] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0073] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0074] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0075] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0076] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0077] It should also be noted that 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 limitation, 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.

[0078] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A path planning method for multi-intelligent transport AGVs, characterized in that, The method includes: In response to receiving a path planning request from a target AGV, based on pre-built map topology data, the starting path point of the target AGV is taken as the first current path point, and a path search is performed on the first current path point until the ending path point is reached: Find all first adjacent reachable path points of the first current path point; For each of the first adjacent reachable path points, the road segment between the first adjacent reachable path point and the first current path point is taken as the first sub-path to be evaluated corresponding to the first adjacent reachable path point. Based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first actual driving cost of the first adjacent reachable path point is determined. For each of the first adjacent reachable path points, the first cumulative cost of the first current path point is added to the first actual driving cost of the first adjacent reachable path point to obtain the first cumulative cost of the first adjacent reachable path point, and the first cumulative cost of the first adjacent reachable path point is added to the first heuristic cost of the first adjacent reachable path point to obtain the first evaluation cost of the first adjacent reachable path point. Store each of the first adjacent reachable path points in the first open list, and delete the first current path point from the first open list; The path point corresponding to the minimum first evaluation cost in the first open list is taken as the new first current path point; In response to the new first current path point being the termination path point, a global planned path from the starting path point to the termination path point is backtracked to enable the target AGV to travel along the global planned path, and a lock flag is added to the global planned path.

2. The method according to claim 1, characterized in that, The determination of the first actual travel cost of the first adjacent reachable path point based on the spatiotemporal overlap information between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario includes: Based on the first basic cost mapping rule, the travel attributes of the first sub-path to be evaluated corresponding to the first adjacent reachable path point are mapped to obtain the first basic travel cost of the first adjacent reachable path point. The travel attributes include at least one of road segment length, road segment width, and road surface condition. In response to the fact that there is no overlap between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the first basic driving cost of the first adjacent reachable path point is taken as the first actual driving cost of the first adjacent reachable path point. In response to the existence of an overlap between the first sub-path to be evaluated corresponding to the first adjacent reachable path point and the planned path information of other AGVs in the preset scenario, based on the driving direction of the target AGV and other AGVs in the preset scenario in the overlap segment, the first conflict penalty cost of the first adjacent reachable path point is determined, and the first basic driving cost of the first adjacent reachable path point is added to the first conflict penalty cost of the first adjacent reachable path point to obtain the first actual driving cost of the first adjacent reachable path point.

3. The method according to claim 2, characterized in that, The determination of the first conflict penalty cost for the first adjacent reachable path point based on the travel direction of the target AGV in the overlapping segment and other AGVs in the preset scenario includes: The first number of times the target AGV and other AGVs in the preset scenario travel in the same direction in the overlapping segment is counted, and the first number is multiplied by the first preset penalty cost to obtain the first same-direction conflict penalty sub-cost of the first adjacent reachable path point; The second number of times the target AGV in the overlapping segment travels in the opposite direction to other AGVs in the preset scenario is counted, and the second number is multiplied by the second preset penalty cost to obtain the first reverse conflict penalty sub-cost of the first adjacent reachable path point; The first conflict penalty cost of the first adjacent reachable path point is obtained by adding the first same-direction conflict penalty cost of the first adjacent reachable path point to the first conflict penalty cost of the first adjacent reachable path point.

4. The method according to claim 1, characterized in that, The method further includes: The driving status of the target AGV is acquired in real time; In response to the target AGV's driving state being locked and blocked, the target AGV is controlled to drive to the nearest temporary avoidance point, and the locking flag of the target AGV's global planned path is released; Based on pre-built map topology data, the nearest temporary avoidance point of the target AGV is used as the second current path point, and a path search is performed on the second current path point until the termination path point is reached: Find all second adjacent reachable path points of the second current path point; For each second adjacent reachable path point, the road segment between the second adjacent reachable path point and the second current path point is taken as the second sub-path to be evaluated corresponding to the second adjacent reachable path point. Based on the spatiotemporal overlap information between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the second actual driving cost of the second adjacent reachable path point is determined. For each of the second adjacent reachable path points, the second cumulative cost of the second current path point is added to the second actual driving cost of the second adjacent reachable path point to obtain the second cumulative cost of the second adjacent reachable path point, and the second cumulative cost of the second adjacent reachable path point is added to the second heuristic cost of the second adjacent reachable path point to obtain the second evaluation cost of the second adjacent reachable path point; Store each of the second adjacent reachable path points into the second open list, and delete the second current path point from the second open list; The path point corresponding to the minimum second evaluation cost in the second open list is taken as the new second current path point; In response to the new second current path point being the termination path point, a valid planned path from the nearest temporary avoidance point to the termination path point is backtracked to enable the target AGV to continue traveling along the valid planned path, and a lock flag is added to the valid planned path.

5. The method according to claim 4, characterized in that, The determination of the second actual travel cost of the second adjacent reachable path point based on the spatiotemporal overlap information between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario includes: Based on the second basic cost mapping rule, the traffic attributes of the second sub-path to be evaluated corresponding to the second adjacent reachable path point are mapped to obtain the second basic driving cost of the second adjacent reachable path point. The traffic attributes include at least one of road segment length, road segment width, and road surface condition. In response to the fact that there is no overlap between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the second basic driving cost of the second adjacent reachable path point is taken as the second actual driving cost of the second adjacent reachable path point. In response to the existence of an overlap between the second sub-path to be evaluated corresponding to the second adjacent reachable path point and the planned path information of other AGVs in the preset scenario, based on the driving direction of the target AGV and other AGVs in the preset scenario in the overlap segment, the second conflict penalty cost of the second adjacent reachable path point is determined, and the second basic driving cost of the second adjacent reachable path point is added to the second conflict penalty cost of the second adjacent reachable path point to obtain the second actual driving cost of the second adjacent reachable path point.

6. The method according to claim 5, characterized in that, The determination of the second conflict penalty cost for the second adjacent reachable path point based on the travel direction of the target AGV in the overlapping segment and other AGVs in the preset scenario includes: The third number of times the target AGV and other AGVs in the preset scenario travel in the same direction in the overlapping segment is counted, and the third number is multiplied by the third preset penalty cost to obtain the second same-direction conflict penalty sub-cost of the second adjacent reachable path point; The fourth number of times the target AGV and other AGVs in the preset scenario travel in opposite directions in the overlapping segment is counted, and the fourth number is multiplied by the fourth preset penalty cost to obtain the second reverse conflict penalty sub-cost of the second adjacent reachable path point; The second conflict penalty cost of the second adjacent reachable path point is obtained by adding the second same-direction conflict penalty cost of the second adjacent reachable path point to the second conflict penalty cost of the second adjacent reachable path point.

7. The method according to claim 4, characterized in that, The method further includes: In response to the presence of a parked AGV within the preset area, based on pre-constructed map topology data, the parking point of the parked AGV is designated as the third current path point, and a path search is performed on the third current path point until a waiting path point is reached: Find all third adjacent reachable path points of the third current path point; For each of the third adjacent reachable path points, the road segment between the third adjacent reachable path point and the third current path point is taken as the third sub-path to be evaluated corresponding to the third adjacent reachable path point. Based on the spatiotemporal overlap information between the third sub-path to be evaluated corresponding to the third adjacent reachable path point and the planned path information of other AGVs in the preset scenario, the third real driving cost of the third adjacent reachable path point is determined. For each of the third adjacent reachable path points, the third cumulative cost of the third current path point is added to the third actual driving cost of the third adjacent reachable path point to obtain the third cumulative cost of the third adjacent reachable path point, and the third cumulative cost of the third adjacent reachable path point is added to the third heuristic cost of the third adjacent reachable path point to obtain the third evaluation cost of the third adjacent reachable path point. Store each of the third adjacent reachable path points in the third open list, and delete the third current path point from the third open list; The path point corresponding to the minimum third evaluation cost in the third open list is taken as the new third current path point; In response to the new third current path point being the waiting path point, a transfer planning path from the starting path point to the waiting path point is backtracked to enable the parked AGV to transfer to the waiting path point along the transfer planning path.

8. The method according to claim 4, characterized in that, The method further includes: In response to other AGVs that cause locking conflicts leaving the locked area or completing their work at the workstation and transferring to the designated parking point, the locking record of the corresponding road segment is automatically cleared, restoring the right-of-way for that road segment and allowing the other AGVs in the yard to pass normally.

9. A multi-intelligent handling AGV path planning device, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method as described in any one of claims 1-8.

10. A computer storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed, they implement the method as described in any one of claims 1-8.

Images