Automatic guided vehicle path planning and scheduling method, device and equipment and storage medium

Abstract

An automated guided vehicle (AGV) path planning and scheduling method, apparatus, equipment, and storage medium are disclosed, relating to the field of intelligent scheduling. In this method, if a time window change event is received, the spatiotemporal influence domain is determined based on the time window change event and a preset spatiotemporal network. By transforming the plan change into a spatiotemporal constraint range, deep integration of the time dimension of production scheduling and path planning is achieved, solving the problem of path and time disconnection. Based on the spatiotemporal influence domain and the initial spatiotemporal trajectory of the pre-planned AGV, the target AGV and its corresponding path segment to be replanned are determined, and the affected target AGVs are screened to avoid invalid calculations for irrelevant AGVs. For the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory. Based on the updated spatiotemporal trajectory, the target AGV is controlled to perform material handling tasks. By recalculating only the affected path segment instead of a full recalculation, the response time is significantly shortened.

Description

Technical Field

[0001] This application relates to the field of intelligent scheduling, specifically to an automated guided vehicle (AGV) path planning and scheduling method, device, equipment, and storage medium. Background Technology

[0002] With the rapid development of intelligent manufacturing technology, industrial manufacturing scenarios are placing higher demands on the automation and intelligence of production logistics. Automated Guided Vehicles (AGVs), as core equipment for material handling, directly impact the production line's operational rhythm through their path planning efficiency. Meanwhile, Advanced Planning and Scheduling (APS) systems are responsible for developing detailed production process plans, specifying the operating time windows for each machine. To achieve efficient production, the AGV scheduling system and the APS system need to work closely together to ensure that materials are delivered to designated machines on time, meeting the needs of lean manufacturing.

[0003] In related technologies, the APS system and the AGV scheduling system are often loosely coupled. After the APS issues a task instruction, the AGV system independently calculates the path. When the production plan is adjusted, the AGV scheduling system usually receives a new task instruction or time requirement and recalculates the travel path of the relevant AGV.

[0004] However, existing technologies have the following technical problems: First, path planning only considers spatial constraints, which leads to a mismatch between AGV arrival time and machine operation time, resulting in either early arrival occupying resources or late arrival causing machine waiting; Second, when the APS plan undergoes minor adjustments, traditional AGV scheduling systems usually use a full recalculation method to update the path, which involves a large amount of calculation and takes a long time, making it impossible to achieve second-level linkage response and difficult to adapt to frequent plan changes on the production site, thus affecting production stability. Summary of the Invention

[0005] This application provides an automated guided vehicle (AGV) path planning and scheduling method, apparatus, equipment, and storage medium, which can solve the technical problems of mismatch between AGV arrival time and machine operation time and large computational load in the prior art.

[0006] In a first aspect, embodiments of this application provide an automated guided vehicle (AGV) path planning and scheduling method, the AGV path planning and scheduling method comprising: If a time window change event is received, the spatiotemporal influence domain is determined based on the time window change event and the preset spatiotemporal network. The preset spatiotemporal network is a network model constructed based on preset workshop geographic information and preset discrete time units. The spatiotemporal influence domain includes spatial range and temporal range. Based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV), the target AGV and the corresponding path segment to be replanned are determined. The initial spatiotemporal trajectory is the path containing position coordinates and time information obtained by the AGV based on the preset spatiotemporal network and production scheduling plan before receiving the time window change event. For the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory, and the target AGV is controlled to perform material handling tasks based on the updated spatiotemporal trajectory.

[0007] In conjunction with the first aspect, in one implementation, determining the target AGV and the corresponding path segment to be replanned includes: The time window change event is analyzed to obtain the time change parameters and associated spatial location, and a spatiotemporal influence domain including the time range and spatial range is constructed, wherein the time range and spatial range are determined based on the time change parameters and associated spatial location; The initial spatiotemporal trajectory is intersected with the spatiotemporal influence domain, and the automated guided vehicle that intersects with the spatiotemporal influence domain is identified as the target AGV. The path segment in the initial spatiotemporal trajectory of the target AGV that is located after the time range of the spatiotemporal influence domain is determined as the path segment to be replanned.

[0008] In conjunction with the first aspect, in one implementation, the step of incrementally replanning the path segment to be replanned in the preset spatiotemporal network for the target AGV to obtain the updated spatiotemporal trajectory includes: Obtain the change time window constraint corresponding to the time window change event; For the target AGV, using the change time window constraint as a constraint condition, the path segment to be replanned is planned in the preset spatiotemporal network to obtain a new planned path segment that does not conflict with the spatiotemporal trajectories of other AGVs in the preset spatiotemporal network. The updated spatiotemporal trajectory is obtained by connecting the other path segments in the initial spatiotemporal trajectory of the target AGV, excluding the path segment to be replanned, with the newly planned path segment.

[0009] In conjunction with the first aspect, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling method further includes: The spatiotemporal trajectories of all AGVs are modeled as geometric bounding boxes, which contain spatial dimension intervals and time dimension intervals; The system detects whether the geometric bounding box corresponding to the newly planned path segment intersects with the geometric bounding boxes of other AGVs in each spatial dimension and time dimension. If there is an intersection, it is determined that there is a spatiotemporal conflict. The priority of each AGV is obtained according to a preset weighted scoring formula, which includes hard constraint level, task urgency, remaining path time margin, and travel time. If a spatiotemporal conflict exists, the AGV with the lowest priority among those involved in the conflict will be the one to give way.

[0010] In conjunction with the first aspect, in one implementation, before determining the target AGV and the corresponding replanning path segment based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the AGV, the method further includes: Obtain the production scheduling plan and extract the machine operation time window corresponding to the production scheduling plan; The machine operation time window is mapped to a hard constraint node in the preset spatiotemporal network; The hard constraint node is set as a necessary node in the path planning, and the time when the AGV arrives at the hard constraint node is constrained within the machine operation time window to form an arrival time window constraint. Based on the arrival time window constraint, the initial spatiotemporal trajectory is generated in the preset spatiotemporal network.

[0011] In conjunction with the first aspect, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling method further includes: A planning field of view is set in the preset spatiotemporal network, and the planning field of view is a planning time range divided along the time axis; The planning field of view is divided into a near-term precise layer and a long-term coarse constraint layer, wherein the time window width of the near-term precise layer is smaller than the time window width of the long-term coarse constraint layer. A four-dimensional spatiotemporal tube is established for each AGV within the near-precision layer, and incremental replanning is performed on the spatiotemporal trajectory located within the near-precision layer based on the four-dimensional spatiotemporal tube; Within the long-term coarse constraint layer, hard constraint nodes are marked and channel resources are reserved, without binding time-space pipes to AGVs; The planning field of view is scrolled forward along the time axis at preset time intervals, and the path segment to be entered into the nearest accurate layer is refined from coarse constraints to accurate four-dimensional spatiotemporal nodes, triggering the next round of path refinement within the planning field of view.

[0012] In conjunction with the first aspect, in one implementation, the time window change event is issued by an Advanced Planning and Scheduling (APS) system, which is used to formulate production scheduling plans and generate the time window change event when the production scheduling plan changes.

[0013] Secondly, embodiments of this application provide an automated guided vehicle (AGV) path planning and scheduling device, the AGV path planning and scheduling device comprising: The first determining module is used to determine the spatiotemporal influence domain based on the time window change event and the preset spatiotemporal network if a time window change event is received. The preset spatiotemporal network is a network model constructed based on preset workshop geographical information and preset discrete time units. The spatiotemporal influence domain includes a spatial range and a temporal range. The second determining module is used to determine the target AGV and the corresponding path segment to be replanned based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV). The planning module is used to perform incremental replanning on the path segment to be replanned in the preset spatiotemporal network for the target AGV, obtain an updated spatiotemporal trajectory, and control the target AGV to perform material handling tasks based on the updated spatiotemporal trajectory.

[0014] Thirdly, embodiments of this application provide an automated guided vehicle (AGV) path planning and scheduling device, which includes a processor, a memory, and an AAV path planning and scheduling program stored in the memory and executable by the processor. When the AAV path planning and scheduling program is executed by the processor, it implements the steps of the AAV path planning and scheduling method as described in the first aspect.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores an automated guided vehicle (AGV) path planning and scheduling program, wherein when the AGV path planning and scheduling program is executed by a processor, it implements the steps of the AGV path planning and scheduling method as described in the first aspect.

[0016] The beneficial effects of the technical solutions provided in this application include: If a time window change event is received, the spatiotemporal influence domain is determined based on the time window change event and the preset spatiotemporal network. By transforming the plan change into a spatiotemporal constraint range, the time dimension of production scheduling and path planning is deeply integrated, solving the problem of path and time disconnection. According to the spatiotemporal influence domain and the initial spatiotemporal trajectory of the pre-planned AGV, the target AGV and the corresponding path segment to be replanned are determined, and the affected target AGVs are screened out to avoid invalid calculations for irrelevant AGVs. For the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory. Based on the updated spatiotemporal trajectory, the target AGV is controlled to perform material handling tasks. By recalculating only the affected path segment instead of recalculating the entire path, the amount of calculation is reduced and the response time is significantly shortened. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating an embodiment of the automated guided vehicle path planning and scheduling method of this application; Figure 2 This is a detailed schematic diagram of step S20 of this application; Figure 3 This is a detailed schematic diagram of step S30 of this application; Figure 4 This is a schematic diagram of the functional modules of an embodiment of the automated guided vehicle path planning and scheduling device of this application; Figure 5 This is a schematic diagram of the hardware structure of the automated guided vehicle path planning and scheduling equipment involved in the embodiments of this application. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0020] In a first aspect, embodiments of this application provide an automated guided vehicle (AGV) path planning and scheduling method.

[0021] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the automated guided vehicle (AGV) path planning and scheduling method of this application. Figure 1As shown, the automated guided vehicle (AGV) path planning and scheduling method includes: Step S10: If a time window change event is received, then based on the time window change event and the preset spatiotemporal network, determine the spatiotemporal influence domain, wherein the preset spatiotemporal network is a network model constructed based on preset workshop geographical information and preset discrete time units, and the spatiotemporal influence domain includes spatial range and time range. In this embodiment, if a time window change event is received from the Advanced Planning and Scheduling (APS) system, the spatiotemporal influence domain is determined based on the time window change event and the preset spatiotemporal network.

[0022] Specifically, the preset spatiotemporal network is a four-dimensional data model constructed based on preset workshop geographic information and preset discrete time units. In this network, nodes not only represent physical locations but also the state of those locations in a specific time segment; edges represent the permission for automated guided vehicles (AGVs) to move from one location to another between adjacent time segments. The specific data structure is defined as follows: the spatiotemporal nodes of the preset spatiotemporal network are represented by a quadruple N=(x,y,z,t_k), where (x,y,z) are the three-dimensional coordinates of the workshop location, t_k is the discrete time segment number (preferably 30 seconds for the mold workshop), and it is accompanied by state attributes S∈{idle, reserved, occupied, hard constraint} and associated attributes {occupied AGV number, associated APS process ID, time window label TW}. Spatiotemporal edges are represented by a quintuple E=(N_i,N_j,Δt_ij,v_ij,TW_ij), where N_i and N_j are the starting and target spatiotemporal nodes of the edge, respectively; Δt_ij is the time span required to move from node N_i to node N_j; v_ij is the movement speed or passage cost parameter; and each spatiotemporal edge is accompanied by a time window label TW_ij, indicating the movement permission of the AGV between adjacent time slices. A spatiotemporal tube is the sequence of spatiotemporal nodes occupied by the AGV and its expanded neighborhood, which is geometrically equivalent to a cuboid stretched along the time axis with the AGV shape as the base. Intersection is determined when the geometric projections of two spatiotemporal tubes overlap at any time slice t_k. By constructing this network, the system expands the static spatial map into a dynamic spatiotemporal resource map.

[0023] The preset workshop geographic information is the digitized physical environment data of the factory, including the location coordinates of passages, workstations, stops and obstacles in the workshop. This data is usually entered or imported by map building tools during the system initialization phase as the spatial basis for path planning.

[0024] The default discrete time unit is to divide the continuous time axis into several time segments, such as one unit per minute, one unit per 30 seconds, or one unit per 10 seconds. The purpose of introducing the discrete time unit is to transform the continuous time dimension into a discrete state that can be processed by a computer so that it can be combined with spatial coordinates.

[0025] The spatiotemporal influence domain refers to a specific area in the preset spatiotemporal network that is affected by the time window change event. This area includes both the affected physical space range (e.g., the area around a machine) and the affected time range (e.g., the time period during which the plan change takes effect). The process of determining the spatiotemporal influence domain is to map the production plan change onto the spatiotemporal network and delineate the spatiotemporal resource areas that need to be focused on so that the path can be adjusted in a targeted manner, rather than performing a full calculation on the entire network.

[0026] Furthermore, in one embodiment, the time window change event is issued by an Advanced Planning and Scheduling (APS) system, which is used to formulate production scheduling plans and generate the time window change event when the production scheduling plan changes.

[0027] In this embodiment, the time window change event is a notification message issued by the production scheduling system due to production plan adjustments (such as urgent order insertion, equipment failure, process delay, etc.). This message carries relevant data on the plan change. Specifically, the APS system monitors the production site status and identifies the triggering conditions for plan changes, including: monitoring events such as urgent order insertion, electrode rework, and short-term machine downtime, and determining whether the plan adjustment exceeds a preset threshold (such as ±5 minutes); generating and publishing the time window change event, including: constructing a data packet containing the change type, affected machine ID, and comparison of the old and new time windows, and sending it to the AGV scheduling system through a message queue.

[0028] Step S20: Based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV), determine the target AGV and the corresponding path segment to be replanned. The initial spatiotemporal trajectory is the path containing position coordinates and time information obtained by the AGV based on the preset spatiotemporal network and production scheduling plan before receiving the time window change event. In this embodiment, the target AGV and the corresponding path segment to be replanned are determined based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the AGV. Specifically, the initial spatiotemporal trajectory refers to the task path that the AGV has planned and stored in the scheduling database before receiving the time window change event. Unlike the traditional path that only contains spatial coordinates, the initial spatiotemporal trajectory records when the AGV is expected to arrive at what position. It is a four-dimensional path data containing position coordinates and time information. It is the comparison benchmark for the system to make incremental adjustments and represents the original scheduling intention before the change occurs. The target AGV refers to an AGV whose initial spatiotemporal trajectory is associated with the spatiotemporal influence domain. For example, its travel route passes through the spatial range of the spatiotemporal influence domain, and its expected arrival time falls within the time range of the spatiotemporal influence domain. By screening target AGVs, the system can identify which vehicles have been actually affected by the production plan change, thereby excluding unaffected vehicles and reducing unnecessary waste of computing resources. The path segment to be replanned refers to the part of the initial spatiotemporal trajectory of the target AGV that needs to be recalculated. The purpose of determining this path segment is to clarify the specific object of incremental replanning, ensure that the system only adjusts the necessary path segments and retains the unaffected path segments, thereby improving scheduling efficiency.

[0029] Step S30: For the target AGV, perform incremental replanning on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory, and control the target AGV to perform material handling tasks based on the updated spatiotemporal trajectory.

[0030] In this embodiment, for the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory. Based on the updated spatiotemporal trajectory, the target AGV is controlled to perform material handling tasks. Specifically, incremental replanning refers to the process of re-searching for feasible paths of the path segment to be replanned in the preset spatiotemporal network, under the premise of satisfying the changed time window constraints. During this process, the system will perform conflict detection on the path segment to be replanned and the spatiotemporal trajectories already occupied by other AGVs to ensure that the newly generated path segment does not interfere with other vehicles in time and space. The process of controlling the target AGV to perform material handling tasks based on the updated spatiotemporal trajectory means that the scheduling system converts the updated spatiotemporal trajectory into specific motion control commands, such as speed commands, direction commands, start and stop commands, and sends them to the on-board controller of the target AGV. The on-board controller drives the AGV to complete material delivery according to the received commands and the new spatiotemporal path.

[0031] In this embodiment, if a time window change event is received, the spatiotemporal influence domain is determined based on the time window change event and the preset spatiotemporal network. By converting the plan change into a spatiotemporal constraint range, the time dimension of production scheduling and path planning is deeply integrated, solving the problem of path and time disconnection. According to the spatiotemporal influence domain and the initial spatiotemporal trajectory of the pre-planned AGV, the target AGV and the corresponding path segment to be replanned are determined, and the affected target AGV is screened out to avoid invalid calculations for irrelevant AGVs. For the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory. Based on the updated spatiotemporal trajectory, the target AGV is controlled to perform material handling tasks. By recalculating only the affected path segment instead of recalculating the entire path, the amount of calculation is reduced and the response time is significantly shortened. Tests show that the overall equipment utilization rate (OEE) increases from about 65% to about 78%, the production capacity increases by about 20%, the AGV on-time arrival rate increases from about 78% in the traditional solution to about 94%, and the average waiting time of the machine decreases from about 12 minutes to about 2 minutes, significantly improving production efficiency.

[0032] Furthermore, in one embodiment, determining the target AGV and the corresponding path segment to be replanned includes: Step S201: Analyze the time window change event to obtain the time change parameters and associated spatial location, and construct a spatiotemporal influence domain that includes the time range and spatial range, wherein the time range and spatial range are determined based on the time change parameters and associated spatial location; In this embodiment, as Figure 2 As shown, Figure 2 This is a detailed schematic diagram of step S20 of this application; The change type identifier, time change parameters (such as time offset + 10 min, processing duration 50 min, original planned time window start time) and associated machine number are extracted from the time window change event, and the corresponding spatial coordinates (x, y, z) (i.e., the associated spatial location) are obtained by querying the preset map database based on the machine number; the original planned time window start time is added to the time offset to obtain the new time start time, and the processing duration is used as the time span to form a closed time range [T_start, T_end]; a preset safety distance (such as half the length of the AGV body) is extended outward from the spatial coordinates to form a spatial coverage area; in the preset spatiotemporal network, all nodes within the intersection of the spatial coverage area and the time range are extracted to form a four-dimensional spatiotemporal bounding box as the spatiotemporal influence domain.

[0033] In this embodiment, by mapping change events to a spatiotemporal influence domain that includes both spatial and temporal ranges, the influence range is accurately defined. Compared with the full recalculation scheme, the number of spatiotemporal nodes involved in a single recalculation is reduced by approximately 85% to 92%, significantly reducing the consumption of computing resources. This enables the present application to support real-time linkage scenarios of 10 automated guided vehicles, 50 docking points, and dozens of time window fine-tuning events per day on ordinary industrial servers. Step S202: Perform intersection detection between the initial spatiotemporal trajectory and the spatiotemporal influence domain, and determine the automated guided vehicle that intersects with the spatiotemporal influence domain as the target AGV; Step S203: The path segment in the initial spatiotemporal trajectory of the target AGV that is located after the time range of the spatiotemporal influence domain is determined as the path segment to be replanned.

[0034] In this embodiment, the initial spatiotemporal trajectory coordinate sequence of all AGVs is traversed to determine whether there are any coordinate points that fall within the four-dimensional bounding box of the spatiotemporal influence domain. If so, the AGV is the target AGV. The end time T_end of the spatiotemporal influence domain is obtained, and the path segment with a timestamp greater than T_end and a status marked as "not executed" in the initial spatiotemporal trajectory is extracted and determined as the path segment to be replanned.

[0035] In this embodiment, intersection detection is used to accurately screen affected AGVs, avoiding invalid calculations for irrelevant vehicles. Only the unexecuted path segments after the affected domain are identified as segments to be replanned, while executed paths and unaffected paths are retained.

[0036] Tests show that the number of automated guided vehicles affected by a single fine-tuning event is controlled within 5% to 20% of the total number of vehicles, and the total delay from the occurrence of the event to the completion of the path adjustment is controlled within 5 seconds, which is about 360 times more efficient than the traditional manual adjustment method (about 30 minutes).

[0037] Further, in one embodiment, the step of incrementally replanning the path segment to be replanned in the preset spatiotemporal network for the target AGV to obtain the updated spatiotemporal trajectory includes: Step S301: Obtain the change time window constraint corresponding to the time window change event; Step S302: For the target AGV, using the change time window constraint as a constraint condition, plan the path segment to be replanned in the preset spatiotemporal network to obtain a new planned path segment that does not conflict with the spatiotemporal trajectories of other AGVs in the preset spatiotemporal network. Step S303: Connect the other path segments in the initial spatiotemporal trajectory of the target AGV, excluding the path segment to be replanned, with the newly planned path segment to obtain the updated spatiotemporal trajectory.

[0038] In this embodiment, as Figure 3 As shown, Figure 3 This is a detailed schematic diagram of step S30 of this application; receiving the time window change event sent by the APS system, parsing it into a set of hard constraint nodes in the preset spatiotemporal network, using the A* search algorithm, searching for the shortest path that satisfies the hard constraint nodes while avoiding other spatiotemporal nodes already occupied by other AGVs, retaining the path segments in the initial spatiotemporal trajectory with timestamps less than or equal to T_end as the front segment, and using the newly planned path segments as the back segment, performing smoothing processing at the connection points to generate a complete trajectory, which is the updated spatiotemporal trajectory.

[0039] Furthermore, in one embodiment, before determining the target AGV and the corresponding path segment to be replanned based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV), the method further includes: Obtain the production scheduling plan and extract the machine operation time window corresponding to the production scheduling plan; The machine operation time window is mapped to a hard constraint node in the preset spatiotemporal network; The hard constraint node is set as a necessary node in the path planning, and the time when the AGV arrives at the hard constraint node is constrained within the machine operation time window to form an arrival time window constraint. Based on the arrival time window constraint, the initial spatiotemporal trajectory is generated in the preset spatiotemporal network.

[0040] In this embodiment, the production scheduling plan of the Advanced Planning and Scheduling (APS) system is read through the API interface, and the start and end times of each process are extracted as the machine operation time window. The latest time that the AGV must arrive is calculated based on the process start time minus the material loading time, which is used as the arrival time window constraint. In the preset spatiotemporal network, path search is performed with the arrival time window constraint as the condition to generate the initial spatiotemporal trajectory containing position coordinates and time information. The step of mapping the machine operation time window to a hard constraint node in the preset spatiotemporal network specifically includes: The unified mapping formula for the AGV arrival time window Tagv_arrival: Tagv_arrival∈[Tmachine_start tload tbuffer tmargin,Tmachine_end tprocess tload tbuffer] The machine operation time window is [Tmachine_start, Tmachine_end], the processing time is tprocess, the AGV loading time is tload, the process buffer time is tbuffer, and the running margin is tmargin.

[0041] This formula is applicable to scenarios such as electrode delivery, mold steel delivery, and batch transfer before and after heat treatment furnace. Only the values ​​of tload and tbuffer need to be adjusted. The mapping result is written into the preset spatiotemporal network in the form of "hard constraint nodes + hard constraint time windows". In a pre-defined spatiotemporal network, nodes that match the machine's location coordinates and whose timestamps fall within the machine's operating time window are identified. These nodes are marked as hard constraints. Constraints are configured in the path search algorithm (such as the A* algorithm or Dijkstra's algorithm). If the planned path does not pass through the hard constraint node, or if the timestamp of the hard constraint node exceeds the machine's operating time window, the cost function of that path becomes infinite. This penalty mechanism forces the planning algorithm to generate only paths that meet the time window requirements. Through the node marking and algorithm constraints, the originally soft time reference is transformed into a hard condition in path planning, forming the arrival time window constraint. Preferably, the AGV arrival time window Tagv_arrival is typically narrower than the machine's operating time window.

[0042] Furthermore, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling method further includes: The spatiotemporal trajectories of all AGVs are modeled as geometric bounding boxes, which contain spatial dimension intervals and time dimension intervals; The system detects whether the geometric bounding box corresponding to the newly planned path segment intersects with the geometric bounding boxes of other AGVs in each spatial dimension and time dimension. If there is an intersection, it is determined that there is a spatiotemporal conflict. The priority of each AGV is obtained according to a preset weighted scoring formula, which includes hard constraint level, task urgency, remaining path time margin, and travel time. If a spatiotemporal conflict exists, the AGV with the lowest priority among those involved in the conflict will be the one to give way.

[0043] In this embodiment, the spatiotemporal trajectory of each AGV is abstracted into a geometric bounding box. Each geometric bounding box corresponds to the spatial dimension value range and the time dimension value range, respectively. A corresponding geometric bounding box is constructed for the new path segment obtained by replanning. The geometric bounding boxes of the other AGVs are traversed. If all spatial dimension ranges and time dimension ranges of the two sets of bounding boxes intersect each other, it is determined that there is a spatiotemporal conflict between the two trajectories. To eliminate the uncertainty in priority determination during conflict avoidance and to ensure the feasibility of this embodiment, the following conflict resolution strategy is adopted: 1. Definition of Automated Guided Vehicle (AGV) Priority: The AGV priority P is defined as a weighted score. P = w1·H + w2·U + w3·(1 / S) + w4·T Wherein, H is the hard constraint level, indicating the urgency of the APS process associated with the current task of the automated guided vehicle, and the value can be selected from 1 to 3. The narrower the hard constraint time window, the higher H is; U is the task urgency, indicating the priority of the task marked by APS, and the value can be selected from 1 to 5; S is the remaining path time margin, indicating the difference between the current available time window width and the shortest path time, in minutes; T is the travel time, indicating the travel time of the automated guided vehicle since its initiation, in minutes; w1, w2, w3 and w4 are preset weight coefficients, and the default weight order is w1>w2>w3>w4 (preferred w1=0.5, w2=0.3, w3=0.15, w4=0.05).

[0044] 2. Avoidance decision logic: The AGV with the lower P value is the one that avoids the obstacle. If the P values ​​are the same, the decision is made according to the lexicographical order of the AGV numbers to avoid oscillation dead loops in symmetrical scenarios.

[0045] 3. Upgraded processing mechanism: When the number of failed attempts to resolve the temporal and spatial transfer reaches the threshold (preferably 3 times), a detour path is taken; if the detour path is still unsolvable, the process is upgraded to the rolling optimization layer to reallocate the task of the automated guided vehicle and an alarm is sent to the APS system that there is a contention for temporal and spatial resources in the current delivery task.

[0046] In this embodiment, incremental replanning and conflict avoidance are combined to connect the preceding and following path segments, ensuring the continuity of the trajectory, greatly reducing deadlock and conflicts, and improving scheduling stability and equipment utilization.

[0047] In another embodiment, for high-frequency scenarios with minor adjustments to the APS time window of ±5~15 minutes, incremental replanning is performed using a "affect domain boundary determination + incremental path correction" algorithm for hot recalculation, specifically including the following four steps: Step 1: Spatiotemporal recalibration. Based on the time window change event ΔTW=(process ID, TWold, TWnew) published by APS, where TWold is the original time window before the change and TWnew is the new time window after the change, locate the set of hard constraint nodes H bound to the process, and replace the time window label of the set of hard constraint nodes H with the new time window TWnew in the preset spatiotemporal network.

[0048] Step 2: Determine the boundary of the influence domain. Using [TWold∪TWnew] as the initial boundary, expand outward with a safety time margin ΔTsafe (preferably 60~180 seconds) to form the influence spacetime box B (i.e., the specific manifestation of the spacetime influence domain). Only AGVs whose spacetime tubes intersect with the influence spacetime box B are added to the recalculation queue Q, while the remaining AGVs keep their original paths unchanged. Through this step, the number of AGVs affected by a single fine-tuning event is controlled within 5%~20% of the total number of AGVs in the vehicle.

[0049] Step 3: Conflict Detection and Slide Propagation. For the AGVs in the recalculation queue Q, an incremental path search is performed, starting from the current node and ending at the new hard constraint time window. If the new path causes a new conflict with the spatiotemporal management of other AGVs, the other AGVs involved in the conflict are recursively added to the recalculation queue Q until the queue stabilizes. This mechanism ensures that local adjustments do not trigger new global conflicts.

[0050] Step 4: Incremental Path Correction. Only the path segments of the AGV in the queue to be recalculated Q after the end time of the time in the spatiotemporal influence domain are replaced; already executed segments are retained. Through this algorithm, the linkage response time is reduced from several seconds to several minutes to the second level (preferably within 5 seconds).

[0051] In this embodiment, the spatiotemporal constraint transmission for triggering local hot recalculation of AGV paths by minor adjustments to the APS plan is achieved through the above four-step algorithm. Compared with the traditional full recalculation scheme, the linkage response time is reduced from several seconds to several minutes to the second level, significantly improving the system's adaptability to frequent on-site plan changes.

[0052] Furthermore, the computational complexity analysis of the above-mentioned influence domain hot recalculation algorithm is as follows: The time complexity of full recalculation is O(N·M), where N is the total number of AGVs and M is the total number of spatiotemporal nodes; the time complexity of influence domain hot recalculation is reduced to O(|Q|·m), where |Q| is the size of the set of affected AGVs in the queue to be recalculated and m is the number of spatiotemporal nodes in the influence spatiotemporal box. In a typical scenario in a mold workshop (10 AGVs, 50 docking points, time window fine-tuning ±5~15 minutes), |Q|≤2, m≈6~12, the complexity is reduced by more than 90% compared to full recalculation. For non-convergent degenerate scenarios: if the queue to be recalculated Q continues to expand after K rounds (preferably K=3) of recursive iteration... If the scope of the new conflict exceeds expectations, it automatically degenerates into "full recalculation within the impact spacetime box" (still less than the full recalculation of the entire vehicle), ensuring the termination of the algorithm. If there is still no feasible solution within this scope, it rolls back to the next rolling cycle for replanning and triggers a reverse alarm to the APS via the event bus. Optionally, when the situation of "hard constraint node is unreachable within the adjusted time window" occurs in the impact domain determination, the AGV scheduling system sends the ΔTW_infeasible event (including process ID and the earliest reachable time) back to the APS via the event bus. The APS then makes compensatory fine-tuning of the time window of the next process, forming a two-way spacetime constraint closed loop between the APS and the AGV scheduling layer.

[0053] Furthermore, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling method further includes: A planning field of view is set in the preset spatiotemporal network, and the planning field of view is a planning time range divided along the time axis; The planning field of view is divided into a near-term precise layer and a long-term coarse constraint layer, wherein the time window width of the near-term precise layer is smaller than the time window width of the long-term coarse constraint layer. A four-dimensional spatiotemporal tube is established for each AGV within the near-precision layer, and incremental replanning is performed on the spatiotemporal trajectory located within the near-precision layer based on the four-dimensional spatiotemporal tube; Within the long-term coarse constraint layer, hard constraint nodes are marked and channel resources are reserved, without binding time-space pipes to AGVs; The planning field of view is scrolled forward along the time axis at preset time intervals, and the path segment to be entered into the nearest accurate layer is refined from coarse constraints to accurate four-dimensional spatiotemporal nodes, triggering the next round of path refinement within the planning field of view.

[0054] In this embodiment, a rolling horizon approach is used for global scheduling, dividing the planning horizon into two sub-layers: a near-term precise layer and a long-term coarse constraint layer. This achieves a balance between long-term stability and short-term real-time response. For all AGVs falling into the near-term precise layer, a dedicated four-dimensional spatiotemporal tube is constructed based on the workshop's three-dimensional coordinates and discrete time, fully recording the space and timing occupied by each AGV within the near-term window. When subsequent APS plans change, incremental replanning is performed on the spatiotemporal trajectories within the near-term layer using only this four-dimensional spatiotemporal tube, without requiring global recalculation of the unrefined long-term paths. The long-term coarse constraint layer only stores the hard constraint nodes corresponding to APS processes, reserving passage resources in a channel-level spatiotemporal corridor, and does not allocate dedicated four-dimensional spatiotemporal tubes to any AGV. If only the time window within the long-term window is slightly adjusted, the system only modifies the time labels of the hard constraint nodes, without triggering AGV path replanning, significantly reducing computational overhead. Specific implementation details are as follows: 1. Two-layer field of view division: Recent Precision Layer: The time window width Wnear is preferably 15 minutes, and the planning granularity is aligned with the time slice of the preset spatiotemporal network (preferably 30 seconds). Within this window, a precise four-dimensional spatiotemporal tube is established for each AGV, and the incremental replanning and deadlock avoidance described in the above embodiments are executed. This layer is used to guide the immediate execution actions of the AGV.

[0055] Long-term coarse constraint layer: The time window width Wfar is preferably extended to the end of the shift (approximately 6-8 hours). APS hard constraint nodes (i.e., the arrival time window lock nodes in the APS plan) are only marked as "channel-level spatiotemporal corridors," without binding spatiotemporal management to specific AGVs; only channel resources are reserved. Long-term uncertainties are thus isolated from near-term precise planning, avoiding full path recalculation caused by fluctuations in the long-term APS plan.

[0056] 2. Rolling Advancement Mechanism: Every preset duration ΔTroll (preferably 3 minutes), the recent window is shifted one step to the right along the time axis. The coarse corridor of the [Wnear, Wnear+ΔTroll] segment that is about to enter the recent layer is refined into a precise four-dimensional spatiotemporal node. A lightweight conflict detection is performed on the newly refined segment. At the same time, the historical spatiotemporal management nodes that have been executed and exited the recent layer are archived and released to maintain stable memory usage.

[0057] 3. Collaboration with the influence domain algorithm: When APS publishes a time window change event, the system first determines whether the change time window falls within the recent precision layer (within Wnear): If so: trigger the determination of the spatiotemporal influence domain in step S10 and the hot recalculation of the influence domain in the above embodiment, and update the accurate spatiotemporal tube in the recent layer.

[0058] If not: Only update the time window label of the corresponding hard constraint node in the long-term coarse constraint layer, without triggering any AGV path recalculation, thereby significantly reducing the computational overhead caused by fluctuations in long-term plans.

[0059] 4. Algorithm skeleton (Plan / Refine / Commit / Roll): Plan (Initial Planning): When the system starts up or shifts change, accurately establish the four-dimensional spatiotemporal management system for all AGVs in transit within Wnear, and establish a coarse corridor within Wfar according to the APS plan.

[0060] Refine: Each time a hot recalculation or rolling advancement is triggered, the path segment newly entering the recent layer is refined from a coarse corridor to a precise four-dimensional node.

[0061] Commit (Execution Submission): Pushes the planned path segments in the recent layer to the AGV control system for execution. After execution, the status of the spatiotemporal management node is updated from "Reserved" to "Occupied".

[0062] Roll (Rolling Progress): Every ΔTroll, the time window is shifted to the right, triggering the next Refine round, and this process is repeated until the shift ends.

[0063] In this embodiment, the long-term planning and short-term execution are decoupled by the rolling optimization layer, which reduces the time spent on a single complete path planning. In the scenario of a mold workshop (preferably 10 AGVs and 50 docking points), the scalability and long-term stability of the system are significantly improved, and the average number of monthly oscillations is reduced. Specifically, in the same mold workshop environment with 10 AGVs, 50 docking points, and approximately 200 delivery tasks per day, the method of this application and the traditional independent AGV scheduling scheme were compared continuously for 3 months, as shown in Table 1, the first group of key indicators comparison table of this application. The data in Table 1 shows that in the scenario of strict APS time window, high frequency of AGV path conflict, and frequent fine-tuning of plan events in the mold workshop, the method of this application has achieved significant improvements over the independent AGV scheduling scheme in core indicators such as conflict avoidance, machine waiting, plan on time, response delay, and computational overhead. Under the single precise layer planning scheme using only influence domain hot recalculation, the method in this application can stably support a workshop scale of 10 AGVs, 50 docking points, and 200 delivery tasks per day on a general industrial server. After superimposing the rolling optimization layer, it can stably support a workshop scale of ≥20 AGVs, ≥100 docking points, and ≥400 delivery tasks per day. Compared with the upper limit of about 5 AGVs that can be supported by the single precise layer planning, it can support a scale expansion of about 4 times. At the same time, the long-range oscillation phenomenon (the same conflict repeatedly occurs in multiple planning cycles) is resolved in advance by the coarse corridor reservation mechanism during rolling advancement, and the average number of monthly oscillations is reduced from about 8 times to ≤1 time, further improving the long-term stability of the scheduling system.

[0064] Table 1 Comparison of Key Indicators in the First Group of This Application

[0065] Furthermore, as shown in Table 2, which is a comparison table of the second set of key indicators in this application, the three indicators in Table 2 reflect the independent contribution of the rolling optimization layer in terms of planning efficiency and system scalability: by decomposing the planning horizon into a near-term accurate layer (W_near=15 minutes) and a long-term coarse constraint layer, the system only needs to refine a small batch of spatiotemporal nodes each time it rolls forward, rather than establishing a full-shift accurate path at once, which fundamentally solves the technical contradiction that it is difficult to balance "planning accuracy" and "response real-time performance".

[0066] Table 2 Comparison of Key Indicators in the Second Group of This Application

[0067] Secondly, embodiments of this application also provide an automated guided vehicle (AGV) path planning and scheduling device.

[0068] In one embodiment, reference is made to Figure 4 , Figure 4 This is a functional module diagram of an embodiment of the automated guided vehicle (AGV) path planning and scheduling device of this application. Figure 4 As shown, the automated guided vehicle (AGV) path planning and scheduling device includes: The first determining module is used to determine the spatiotemporal influence domain based on the time window change event and the preset spatiotemporal network if a time window change event is received. The preset spatiotemporal network is a network model constructed based on preset workshop geographical information and preset discrete time units. The spatiotemporal influence domain includes a spatial range and a temporal range. The second determining module is used to determine the target AGV and the corresponding path segment to be replanned based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV). The planning module is used to perform incremental replanning on the path segment to be replanned in the preset spatiotemporal network for the target AGV, obtain an updated spatiotemporal trajectory, and control the target AGV to perform material handling tasks based on the updated spatiotemporal trajectory.

[0069] Furthermore, in one embodiment, the second determining module is used to: The time window change event is analyzed to obtain the time change parameters and associated spatial location, and a spatiotemporal influence domain including the time range and spatial range is constructed, wherein the time range and spatial range are determined based on the time change parameters and associated spatial location; The initial spatiotemporal trajectory is intersected with the spatiotemporal influence domain, and the automated guided vehicle that intersects with the spatiotemporal influence domain is identified as the target AGV. The path segment in the initial spatiotemporal trajectory of the target AGV that is located after the time range of the spatiotemporal influence domain is determined as the path segment to be replanned.

[0070] Furthermore, in one embodiment, the planning module is used for: Obtain the change time window constraint corresponding to the time window change event; For the target AGV, using the change time window constraint as a constraint condition, the path segment to be replanned is planned in the preset spatiotemporal network to obtain a new planned path segment that does not conflict with the spatiotemporal trajectories of other AGVs in the preset spatiotemporal network. The updated spatiotemporal trajectory is obtained by connecting the other path segments in the initial spatiotemporal trajectory of the target AGV, excluding the path segment to be replanned, with the newly planned path segment.

[0071] Furthermore, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling device further includes a modeling module, used for: The spatiotemporal trajectories of all AGVs are modeled as geometric bounding boxes, which contain spatial dimension intervals and time dimension intervals; The system detects whether the geometric bounding box corresponding to the newly planned path segment intersects with the geometric bounding boxes of other AGVs in each spatial dimension and time dimension. If there is an intersection, it is determined that there is a spatiotemporal conflict. The priority of each AGV is obtained according to a preset weighted scoring formula, which includes hard constraint level, task urgency, remaining path time margin, and travel time. If a spatiotemporal conflict exists, the AGV with the lowest priority among those involved in the conflict will be the one to give way.

[0072] Furthermore, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling device further includes an acquisition module, used for: Obtain the production scheduling plan and extract the machine operation time window corresponding to the production scheduling plan; The machine operation time window is mapped to a hard constraint node in the preset spatiotemporal network; The hard constraint node is set as a necessary node in the path planning, and the time when the AGV arrives at the hard constraint node is constrained within the machine operation time window to form an arrival time window constraint. Based on the arrival time window constraint, the initial spatiotemporal trajectory is generated in the preset spatiotemporal network.

[0073] Furthermore, in one embodiment, the automated guided vehicle (AGV) path planning and scheduling device further includes a partitioning module, used for: A planning field of view is set in the preset spatiotemporal network, and the planning field of view is a planning time range divided along the time axis; The planning field of view is divided into a near-term precise layer and a long-term coarse constraint layer, wherein the time window width of the near-term precise layer is smaller than the time window width of the long-term coarse constraint layer. A four-dimensional spatiotemporal tube is established for each AGV within the near-precision layer, and incremental replanning is performed on the spatiotemporal trajectory located within the near-precision layer based on the four-dimensional spatiotemporal tube; Within the long-term coarse constraint layer, hard constraint nodes are marked and channel resources are reserved, without binding time-space pipes to AGVs; The planning field of view is scrolled forward along the time axis at preset time intervals, and the path segment to be entered into the nearest accurate layer is refined from coarse constraints to accurate four-dimensional spatiotemporal nodes, triggering the next round of path refinement within the planning field of view.

[0074] Furthermore, in one embodiment, the time window change event is issued by an Advanced Planning and Scheduling (APS) system, which is used to formulate production scheduling plans and generate the time window change event when the production scheduling plan changes.

[0075] The functions of each module in the above-mentioned automated guided vehicle path planning and scheduling device correspond to the steps in the above-mentioned automated guided vehicle path planning and scheduling method embodiment, and their functions and implementation processes will not be described in detail here.

[0076] Thirdly, embodiments of this application provide an automated guided vehicle (AGV) path planning and scheduling device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0077] Reference Figure 5 , Figure 5This is a schematic diagram of the hardware structure of the automated guided vehicle (AGV) path planning and scheduling device involved in the embodiments of this application. In the embodiments of this application, the AGV path planning and scheduling device may include a processor, a memory, a communication interface, and a communication bus.

[0078] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0079] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces used for interconnecting components within the automated guided vehicle (AGV) path planning and scheduling equipment, as well as interfaces used for interconnecting the AGV path planning and scheduling equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0080] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0081] The processor can be a general-purpose processor, which can call the automated guided vehicle (AGV) path planning and scheduling program stored in the memory and execute the AGV path planning and scheduling method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the AGV path planning and scheduling program is called can be referred to in the various embodiments of the AGV path planning and scheduling method of this application, and will not be repeated here.

[0082] Those skilled in the art will understand that Figure 5 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0083] Fourthly, embodiments of this application also provide a computer-readable storage medium.

[0084] The present application has an automated guided vehicle (AGV) path planning and scheduling program stored on a computer-readable storage medium, wherein when the AGV path planning and scheduling program is executed by a processor, it implements the steps of the AGV path planning and scheduling method described above.

[0085] The method implemented when the automated guided vehicle path planning and scheduling program is executed can be referred to in various embodiments of the automated guided vehicle path planning and scheduling method of this application, and will not be repeated here.

[0086] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0087] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0088] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0089] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0090] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0091] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0092] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for path planning and scheduling of automated guided vehicles, characterized in that, The automated guided vehicle (AGV) path planning and scheduling method includes: If a time window change event is received, the spatiotemporal influence domain is determined based on the time window change event and the preset spatiotemporal network. The preset spatiotemporal network is a network model constructed based on preset workshop geographic information and preset discrete time units. The spatiotemporal influence domain includes spatial range and temporal range. Based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV), the target AGV and the corresponding path segment to be replanned are determined. The initial spatiotemporal trajectory is the path containing position coordinates and time information obtained by the AGV based on the preset spatiotemporal network and production scheduling plan before receiving the time window change event. For the target AGV, incremental replanning is performed on the path segment to be replanned in the preset spatiotemporal network to obtain an updated spatiotemporal trajectory, and the target AGV is controlled to perform material handling tasks based on the updated spatiotemporal trajectory.

2. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 1, characterized in that, The determination of the target AGV and the corresponding path segment to be replanned for the target AGV includes: The time window change event is analyzed to obtain the time change parameters and associated spatial location, and a spatiotemporal influence domain including the time range and spatial range is constructed, wherein the time range and spatial range are determined based on the time change parameters and associated spatial location; The initial spatiotemporal trajectory is intersected with the spatiotemporal influence domain, and the automated guided vehicle that intersects with the spatiotemporal influence domain is identified as the target AGV. The path segment in the initial spatiotemporal trajectory of the target AGV that is located after the time range of the spatiotemporal influence domain is determined as the path segment to be replanned.

3. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 1, characterized in that, The step of incrementally replanning the path segment to be replanned in the preset spatiotemporal network for the target AGV to obtain the updated spatiotemporal trajectory includes: Obtain the change time window constraint corresponding to the time window change event; For the target AGV, using the change time window constraint as a constraint condition, the path segment to be replanned is planned in the preset spatiotemporal network to obtain a new planned path segment that does not conflict with the spatiotemporal trajectories of other AGVs in the preset spatiotemporal network. The updated spatiotemporal trajectory is obtained by connecting the other path segments in the initial spatiotemporal trajectory of the target AGV, excluding the path segment to be replanned, with the newly planned path segment.

4. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 3, characterized in that, The automated guided vehicle path planning and scheduling method also includes: The spatiotemporal trajectories of all AGVs are modeled as geometric bounding boxes, which contain spatial dimension intervals and time dimension intervals; The system detects whether the geometric bounding box corresponding to the newly planned path segment intersects with the geometric bounding boxes of other AGVs in each spatial dimension and time dimension. If there is an intersection, it is determined that there is a spatiotemporal conflict. The priority of each AGV is obtained according to a preset weighted scoring formula, which includes hard constraint level, task urgency, remaining path time margin, and travel time. If a spatiotemporal conflict exists, the AGV with the lowest priority among those involved in the conflict will be the one to give way.

5. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 1, characterized in that, Before determining the target AGV and the corresponding path segment to be replanned based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the AGV, the following steps are also included: Obtain the production scheduling plan and extract the machine operation time window corresponding to the production scheduling plan; The machine operation time window is mapped to a hard constraint node in the preset spatiotemporal network; The hard constraint node is set as a necessary node in the path planning, and the time when the AGV arrives at the hard constraint node is constrained within the machine operation time window to form an arrival time window constraint. Based on the arrival time window constraint, the initial spatiotemporal trajectory is generated in the preset spatiotemporal network.

6. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 1, characterized in that, The automated guided vehicle path planning and scheduling method also includes: A planning field of view is set in the preset spatiotemporal network, and the planning field of view is a planning time range divided along the time axis; The planning field of view is divided into a near-term precise layer and a long-term coarse constraint layer, wherein the time window width of the near-term precise layer is smaller than the time window width of the long-term coarse constraint layer. A four-dimensional spatiotemporal tube is established for each AGV within the near-precision layer, and incremental replanning is performed on the spatiotemporal trajectory located within the near-precision layer based on the four-dimensional spatiotemporal tube; Within the long-term coarse constraint layer, hard constraint nodes are marked and channel resources are reserved, without binding time-space pipes to AGVs; The planning field of view is scrolled forward along the time axis at preset time intervals, and the path segment to be entered into the nearest accurate layer is refined from coarse constraints to accurate four-dimensional spatiotemporal nodes, triggering the next round of path refinement within the planning field of view.

7. The automated guided vehicle (AGV) path planning and scheduling method as described in claim 1, characterized in that, The time window change event is published by the Advanced Planning and Scheduling (APS) system, which is used to formulate production scheduling plans and generates the time window change event when the production scheduling plan changes.

8. An automated guided vehicle (AGV) path planning and scheduling device, characterized in that, The automated guided vehicle (AGV) path planning and scheduling device includes: The first determining module is used to determine the spatiotemporal influence domain based on the time window change event and the preset spatiotemporal network if a time window change event is received. The preset spatiotemporal network is a network model constructed based on preset workshop geographical information and preset discrete time units. The spatiotemporal influence domain includes a spatial range and a temporal range. The second determining module is used to determine the target AGV and the corresponding path segment to be replanned based on the spatiotemporal influence domain and the pre-planned initial spatiotemporal trajectory of the automated guided vehicle (AGV). The planning module is used to perform incremental replanning on the path segment to be replanned in the preset spatiotemporal network for the target AGV, obtain an updated spatiotemporal trajectory, and control the target AGV to perform material handling tasks based on the updated spatiotemporal trajectory.

9. An automated guided vehicle (AGV) path planning and scheduling device, characterized in that, The automated guided vehicle (AGV) path planning and scheduling device includes a processor, a memory, and an AGV path planning and scheduling program stored in the memory and executable by the processor, wherein when the AGV path planning and scheduling program is executed by the processor, it implements the steps of the AGV path planning and scheduling method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an automated guided vehicle (AGV) path planning and scheduling program, wherein when the AGV path planning and scheduling program is executed by a processor, it implements the steps of the AGV path planning and scheduling method as described in any one of claims 1 to 7.

Images