Scheduling optimization method and system for collaborative operation of multiple AMR robots and medium

Abstract

The invention provides a multi-AMR robot collaborative operation scheduling optimization method and system and a medium. The method comprises the steps that a collaborative operation scheduling map is constructed based on environment information; allocating a corresponding first target task based on the first state information of the AMR robot and the first task information of each to-be-executed task; planning an initial operation route of the AMR robot based on the collaborative operation scheduling map in combination with the second state information of the AMR robot and the second task information of the first target task; performing path conflict detection optimization based on the initial operation path of the AMR robot to obtain a first target operation path; and based on the real-time state information of the AMR robot in the process of executing the first target task, optimizing the first target operation route and the first target task of the AMR robot, and determining a second target operation route and a second target task of the AMR robot. According to the method, the efficiency and the stability of collaborative operation of the multiple AMR robots are improved.

Description

Technical Field

[0001] This invention relates to the field of robotics, and in particular to a scheduling optimization method, system, and medium for collaborative operations of multiple AMR robots. Background Technology

[0002] In many fields such as modern industrial production and logistics, as task complexity increases and efficiency requirements rise, a single robot often struggles to meet actual needs. For example, in large warehouse logistics scenarios, it is necessary to move a large number of goods of different types and storage locations and accurately deliver them to various designated locations. This process involves complex path planning, task allocation, and time management issues. Therefore, multi-robot collaborative operation methods have emerged, where multiple robots cooperate to complete complex tasks.

[0003] However, existing robot collaborative operation methods are not intelligent or efficient enough in handling path conflicts when multiple robots perform tasks simultaneously. For example, after multi-robot path planning is completed, the robots determine the order in which they pass through key locations such as intersections based on the relative time attributes of path points. But in actual operation, if a robot encounters an obstacle midway, speed planning needs to be re-executed. Moreover, when there are many robots, due to factors such as communication delays in the actual environment, it is difficult to guarantee the synchronization of start signals. At the same time, if only the time sequence is used as the operational constraint, when robots have different speeds and are rotating in place, even if multi-robot path planning is completed, collisions cannot be effectively avoided. Furthermore, the current operational constraints also have the possibility of deadlock (although the possibility of deadlock is small in reality, it still poses a risk), which greatly affects the efficiency and stability of robot collaborative operations. Summary of the Invention

[0004] This invention provides a scheduling optimization method, system, and medium for multi-AMR robot collaborative operations, in order to improve the efficiency and stability of multi-AMR robot collaborative operations.

[0005] In a first aspect, the present invention provides a scheduling optimization method for collaborative operations of multiple AMR robots, comprising:

[0006] A collaborative operation scheduling map is constructed based on the collected environmental information of the current work area;

[0007] Based on the first state information of each AMR robot and the first task information of each task to be executed, a corresponding first target task is assigned to each AMR robot.

[0008] Based on the collaborative operation scheduling map, combined with the second state information of each AMR robot and the second task information of the first target task, the initial operation route of each AMR robot is planned.

[0009] Based on the initial working route of each AMR robot, path conflict detection and optimization are performed to obtain the first target working route of each AMR robot;

[0010] Based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target operation route, the first target operation route and the first target task of each AMR robot are optimized to determine the second target operation route and the second target task of each AMR robot.

[0011] Secondly, the present invention also provides a scheduling optimization system for multi-AMR robot cooperative operations, applied to the scheduling optimization method for multi-AMR robot cooperative operations as described in the first aspect; the scheduling optimization system for multi-AMR robot cooperative operations includes:

[0012] The scheduling map construction module is used to build a collaborative operation scheduling map based on the collected environmental information of the current work area;

[0013] The task allocation module is used to allocate a corresponding first target task to each AMR robot based on the first state information of each AMR robot and the first task information of each task to be executed.

[0014] The operation route planning module is used to plan the initial operation route of each AMR robot based on the collaborative operation scheduling map and the second state information of each AMR robot and the second task information of the first target task.

[0015] The operation route optimization module is used to perform path conflict detection and optimization based on the initial operation route of each AMR robot, so as to obtain the first target operation route of each AMR robot.

[0016] The job scheduling optimization module is used to optimize the first target job route and first target task of each AMR robot based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target job route, and to determine the second target job route and second target task of each AMR robot.

[0017] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the scheduling optimization method for collaborative operation of multiple AMR robots as described above.

[0018] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the scheduling optimization method for collaborative operation of multiple AMR robots as described above.

[0019] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the scheduling optimization method for collaborative operation of multiple AMR robots as described above.

[0020] The scheduling optimization method for multi-AMR robot collaborative operations provided in this invention performs path conflict detection and optimization on the initial operation route of each AMR robot. When a path conflict is detected, the initial operation route with the conflict can be optimized to resolve the conflict in a timely and effective manner, avoiding robot collisions or blockages and improving the efficiency of multi-AMR robot collaborative operations. On the other hand, through a dynamic task route adjustment mechanism, tasks can be reallocated and paths replanned in a timely manner when abnormal situations such as robot malfunctions occur, ensuring the continuity and efficiency of the entire operation process, guaranteeing the efficient and stable operation of multi-AMR robot collaborative operations, and improving the reliability of multi-AMR robot collaborative operations. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the scheduling optimization method for collaborative operation of multiple AMR robots provided in an embodiment of the present invention;

[0022] Figure 2 This is a schematic diagram of the scheduling optimization system for collaborative operation of multiple AMR robots provided in an embodiment of the present invention;

[0023] Figure 3 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] See Figure 1 , Figure 1 This is a flowchart illustrating the scheduling optimization method for multi-AMR robot collaborative operations provided by the present invention. In this embodiment of the invention, the execution entity of the scheduling optimization method for multi-AMR robot collaborative operations is the job scheduling system. Therefore, the scheduling optimization method for multi-AMR robot collaborative operations includes:

[0026] Step 10: Construct a collaborative operation scheduling map based on the collected environmental information of the current work area.

[0027] Optionally, the job scheduling system collects environmental information of the current work area, including the location, shape and size of obstacles, as well as the boundaries and characteristics of passable areas. The environmental information is preprocessed, including data cleaning, format standardization and coordinate calibration.

[0028] Furthermore, the operation scheduling system models obstacle information, generating three-dimensional or two-dimensional obstacle models based on their location, shape, and size, and accurately marking them at corresponding coordinates on the map. For passable areas, the operation scheduling system divides them into different types of passable areas based on their boundaries and characteristics, such as ground material, slope, and passage width, such as main roads, secondary roads, and operation area passages, and clearly identifies the attribute parameters of each area on the map.

[0029] Furthermore, the job scheduling system integrates obstacle models and passable area information to generate a complete collaborative job scheduling map, as detailed in steps 101 to 105.

[0030] Step 20: Based on the first state information of each AMR robot and the first task information of each task to be executed, assign a corresponding first target task to each AMR robot.

[0031] Furthermore, the job scheduling system collects the initial state information of all AMR robots, focusing on extracting the initial position coordinates of each AMR robot. Simultaneously, the job scheduling system aggregates the initial task information of all tasks to be executed, clarifying the starting position and priority of each task.

[0032] Furthermore, the job scheduling system establishes a task allocation model that prioritizes task priority, ensuring the allocation of high-priority tasks. Simultaneously, considering the distance between the AMR robot's initial position and the task's starting position, the model allocates tasks to AMR robots with closer initial positions to minimize robot movement distance and time costs. During allocation, it ensures that each AMR robot receives a reasonable number of tasks, avoiding situations where some robots have too many tasks while others are idle. Finally, based on the calculation results of the task allocation model, the job scheduling system determines the corresponding primary target task for each AMR robot and sends the task information to the appropriate robot.

[0033] In one embodiment, there are three AMR robots with initial positions of AMR1 at (5, 5), AMR2 at (25, 25), and AMR3 at (75, 75). There are four tasks to be executed: Task A is high priority, starting at (15, 15); Task B is medium priority, starting at (35, 35); Task C is medium priority, starting at (65, 65); and Task D is low priority, starting at (45, 45). Since Task A has the highest priority, it is assigned a robot first. The distance from each robot's initial position to Task A's starting position is calculated. AMR1 is the closest, so Task A is assigned to AMR1. Next, the medium priority tasks are processed. Task B starts at (35, 35), and AMR2's initial position (25, 25) is close, so Task B is assigned to AMR2. Task C starts at (65, 65), and AMR3's initial position (75, 75) is close, so Task C is assigned to AMR3. Finally, low-priority task D is processed. Of the remaining robots, AMR2 and AMR3 are relatively close to the starting position of task D (45, 45). Considering the number of tasks, AMR2 currently only has one task, so task D is assigned to AMR2. The final allocation result is: AMR1's primary objective task is task A, AMR2's primary objective tasks are tasks B and D, and AMR3's primary objective task is task C.

[0034] Step 30: Based on the collaborative operation scheduling map and combined with the second state information of each AMR robot and the second task information of the first target task, plan the initial operation route of each AMR robot.

[0035] Furthermore, the job scheduling system, based on a collaborative job scheduling map, acquires information on passable areas and obstacle distribution within the work zone. It also collects secondary state information for each AMR robot, including its body width, maximum speed, and battery level. These parameters influence route selection; for example, robots with wider bodies cannot pass through narrow passages, and robots with lower battery levels need to prioritize routes that include charging points. Simultaneously, the job scheduling system defines secondary task information for each primary objective task: the task's starting position, target position, and cargo weight. Cargo weight can affect the robot's speed and energy consumption.

[0036] Furthermore, the job scheduling system comprehensively considers the passable area, obstacles, robot parameters and task information to plan an initial job route for each AMR robot from the initial position to the starting position of the first target task, and then to the target position of the task, as described in steps 301 to 304.

[0037] Step 40: Path conflict detection and optimization are performed based on the initial working route of each AMR robot to obtain the first target working route of each AMR robot.

[0038] Furthermore, the job scheduling system acquires the initial work routes of all AMR robots, decomposing each route into a series of continuous coordinate points and corresponding time nodes, clarifying the position of each robot at different times. Further, the job scheduling system establishes a conflict detection model to compare and analyze the routes of different robots, detecting path conflicts, including spatial conflicts (two or more robots appearing in the same location at the same time or being too close) and temporal conflicts (robots overlapping in time when passing through the same passage or intersection). For detected conflicts, the job scheduling system optimizes the route using conflict resolution strategies based on the severity of the conflict and the priority of the robot's task, such as adjusting the robot's departure time, changing the travel route of some sections, and setting avoidance points. During the optimization process, it must be ensured that the adjusted route still meets the robot's passage requirements and task time constraints. After multiple detections and optimizations, until all path conflicts are resolved, the first target work route for each AMR robot is determined, as detailed in steps 401 to 404.

[0039] Step 50: Based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target operation route, optimize the first target operation route and the first target task of each AMR robot, and determine the second target operation route and the second target task of each AMR robot.

[0040] Furthermore, the job scheduling system receives real-time status information of each AMR robot during the execution of the first target task, including current position coordinates, current moving speed, current remaining battery power, current load weight, and real-time fault codes.

[0041] Furthermore, the job scheduling system analyzes real-time status information to determine whether the robot is traveling normally along the primary target route and whether there are issues such as low battery power, abnormal movement speed, excessive load, or malfunction. When the robot's remaining battery power is found to be below a threshold, the job scheduling system needs to plan a temporary charging route and adjust subsequent tasks. When a robot malfunctions, the job scheduling system needs to reassign its tasks to other normal robots. When a robot deviates from its route due to obstacles or other reasons, the job scheduling system needs to replan its route. Simultaneously, the job scheduling system evaluates the primary target task based on real-time task execution and new task requirements. If a new high-priority task appears or the priority of an existing task changes, the task needs to be reassigned. Finally, based on the analysis and evaluation results, the job scheduling system optimizes the primary target route and primary target task for each AMR robot, determines the secondary target route and secondary target task, and sends them to the corresponding robots, as detailed in steps 501 to 504.

[0042] This invention optimizes the initial work route of each AMR robot by detecting and optimizing the path conflict. When a path conflict is detected, the initial work route with the conflict can be optimized to resolve the conflict in a timely and effective manner, avoiding robot collisions or blockages and improving the efficiency of multi-AMR robot collaborative operations. On the other hand, through a dynamic task route adjustment mechanism, tasks can be reallocated and paths replanned in a timely manner when abnormal situations such as robot malfunctions occur, ensuring the continuity and efficiency of the entire work process, guaranteeing the efficient and stable operation of multi-AMR robot collaborative operations, and improving the reliability of multi-AMR robot collaborative operations.

[0043] Furthermore, the processes in steps 101 to 105 include:

[0044] Step 101: Transform the position, shape and size of the obstacle to obtain the three-dimensional position coordinates and feature parameters of the three-dimensional geometry of the obstacle in the three-dimensional coordinate system. Transform the boundary and features of the passable area to obtain the closed boundary surface parameters of the passable area in three-dimensional space.

[0045] Optionally, for obstacles, the operation scheduling system converts their planar position information into three-dimensional position coordinates in a three-dimensional coordinate system. Typically, a three-dimensional rectangular coordinate system with X, Y, and Z axes is established with a fixed point on the ground of the operation area as the origin. The X and Y axes represent the horizontal direction, and the Z axis represents the vertical direction.

[0046] Furthermore, the job scheduling system converts the shape (such as cuboid, cylinder, irregular shape, etc.) and size parameters (such as length, width, height, diameter, etc.) of obstacles into the characteristic parameters of three-dimensional geometry, such as the length, width, and height parameters of a cuboid, and the base diameter and height parameters of a cylinder.

[0047] For passable areas, the operation scheduling system converts their boundary information into closed boundary surface parameters in three-dimensional space. Through surface equations or boundary point sets, it accurately describes the extent of the passable area in three-dimensional space, including vertical characteristic parameters such as ground height changes and ceiling height restrictions.

[0048] In one embodiment, in the large warehouse operation area, the collected obstacle information includes: a cuboid shelf located 10 meters from the east wall and 20 meters from the south wall, with a length of 5 meters, a width of 2 meters, and a height of 3 meters; and a cylindrical column located 20 meters from the east wall and 30 meters from the south wall, with a base diameter of 1 meter and a height of 4 meters. The boundary of the passable area is the warehouse wall, with the ground ranging from 0 meters to 100 meters from the east wall and from 0 meters to 100 meters from the south wall, a ground height of 0 meters, and a ceiling height of 5 meters. The operation scheduling system establishes a three-dimensional coordinate system with the southeast corner of the warehouse as the origin (0, 0, 0), eastward as the positive X-axis, northward as the positive Y-axis, and upward as the positive Z-axis. The position of the cuboid shelf is converted to three-dimensional coordinates (10, 20, 0), and its three-dimensional geometric features are a length of 5 meters (X-axis), a width of 2 meters (Y-axis), and a height of 3 meters (Z-axis). The position of the cylindrical column is converted to three-dimensional coordinates (20, 30, 0), and its three-dimensional geometric features are a base diameter of 1 meter and a height of 4 meters. The closed boundary surface parameters of the passable area are: ground surface Z = 0 (X from 0 to 100, Y from 0 to 100), ceiling surface Z = 5 (X from 0 to 100, Y from 0 to 100), east boundary surface X = 0 (Y from 0 to 100, Z from 0 to 5), west boundary surface X = 100 (Y from 0 to 100, Z from 0 to 5), south boundary surface Y = 0 (X from 0 to 100, Z from 0 to 5), and north boundary surface Y = 100 (X from 0 to 100, Z from 0 to 5).

[0049] Step 102: Determine the spatial occlusion area of ​​the obstacle based on the feature parameters, and construct the topology of the passable area based on the spatial occlusion area and the closed boundary surface parameters.

[0050] Furthermore, based on the obstacle's three-dimensional geometric feature parameters obtained in step 101, the job scheduling system determines the spatial occlusion area formed by each obstacle in three-dimensional space through spatial geometric calculations, that is, the space occupied by the obstacle itself and the spatial range that cannot be passed due to the existence of the obstacle.

[0051] Furthermore, the job scheduling system, combining the closed boundary surface parameters of the spatial occlusion area and the passable area, divides the passable area into multiple independent passable sub-regions. For each sub-region, the system analyzes and determines the channels connecting them, measuring parameters such as width (minimum horizontal distance), height (minimum vertical distance), and length (distance between the two ends of the channel). Finally, using the passable sub-regions as nodes and the connecting channels as edges, the system constructs a passable area topology containing node attributes (sub-region range, characteristics) and edge attributes (channel width, height, length).

[0052] Continuing with the embodiment of step 101, the spatial obstruction area of ​​the cuboid shelf (10, 20, 0) is a cuboid space with X from 10 to 15, Y from 20 to 22, and Z from 0 to 3; the spatial obstruction area of ​​the cylindrical column (20, 30, 0) is a cylindrical space with (20, 30, 0) as the center of the bottom surface, a bottom radius of 0.5 meters, and a height of 4 meters. The operation scheduling system, combined with the closed boundary surface parameters of the passable area, divides the passable area into three passable sub-areas: sub-area A has X from 0 to 10, Y from 0 to 100, and Z from 0 to 5; sub-area B has X from 15 to 20, Y from 0 to 20 and X from 15 to 20, Y from 22 to 100, and Z from 0 to 5; sub-area C has X from 20.5 to 100, Y from 0 to 100, and Z from 0 to 5 (avoiding the obstruction area of ​​the cylindrical column). The channels connecting sub-regions A and B are located at x = 10 to 15, y = 0 to 20 and x = 10 to 15, y = 22 to 100, respectively. These channels have a width of 5 meters (X-axis direction), a height of 5 meters, and a length of 80 meters (Y-axis direction, the sum of the 0 to 20 and 22 to 100 segments). The channels connecting sub-regions B and C are located at x = 20 to 20.5, y = 0 to 100, with a width of 0.5 meters, a height of 5 meters, and a length of 100 meters. In this constructed topology, the nodes are sub-regions A, B, and C, and the edges are the AB channel and the BC channel, with edge attributes of 5 meters width, 5 meters height, and 80 meters length, and 0.5 meters width, 5 meters height, and 100 meters length, respectively.

[0053] Step 103: Based on the topology, match the attribute parameters and traffic constraints of each work unit to obtain the target traffic area of ​​each work unit.

[0054] Furthermore, the job scheduling system collects attribute parameters for each job unit, including the job unit's working width (maximum horizontal dimension), working height (maximum vertical dimension), working length (maximum dimension in the direction of travel), and minimum turning radius. Simultaneously, it defines traffic constraints, including unit size constraints requiring that the job unit's working width not exceed the passage width, working height not exceed the passage height, and that its working length and minimum turning radius adapt to the passage's spatial layout; and passage size constraints requiring that the passage's width, height, and length meet the safe passage requirements of the job unit, preventing collisions with passage boundaries or obstacles. Further, the job scheduling system matches the attribute parameters of each job unit with the size parameters of passable sub-regions and connecting passages based on the topology, selecting passable areas that meet all traffic constraints as the target passable area for that job unit.

[0055] In one embodiment, assume there are two work units. Work unit 1 has the following attribute parameters: work width 1.2 meters, work height 1.5 meters, work length 2 meters, and minimum turning radius 1 meter. Work unit 2 has the following attribute parameters: work width 0.6 meters, work height 1.2 meters, work length 1.5 meters, and minimum turning radius 0.5 meters. The passage constraints are unit size constraints (work width ≤ passage width, work height ≤ passage height) and passage size constraints (passage width ≥ work width + 0.2 meters safety distance, passage height ≥ work height + 0.2 meters safety distance). The work scheduling system matches the attribute parameters of work unit 1 with the passage parameters in the topology: the AB passage width of 5 meters ≥ 1.2 + 0.2 = 1.4 meters, and the height of 5 meters ≥ 1.5 + 0.2 = 1.7 meters, satisfying the constraints; the BC passage width of 0.5 meters < 1.2 + 0.2 = 1.4 meters, not satisfying the constraints. Therefore, the target passage area for work unit 1 is sub-region A, sub-region B, and the AB passage. For work unit 2, the width of channel AB is 5 meters ≥ 0.6 + 0.2 = 0.8 meters, and the height is 5 meters ≥ 1.2 + 0.2 = 1.4 meters. After adjusting the width of channel BC to 1 meter, 1 meter ≥ 0.6 + 0.2 = 0.8 meters, satisfying the constraints. At this point, the attribute parameters of work unit 2 match both channel AB and channel BC. Therefore, its target passage area is sub-region A, sub-region B, sub-region C, and channels AB and BC.

[0056] Step 104: Perform region overlap detection based on the target passage areas of any two work units to obtain overlapping conflict areas, and perform channel intersection detection based on the connecting channel of the target passage areas of any two work units to obtain intersection conflict areas.

[0057] Furthermore, the job scheduling system performs area overlap detection for the target passage area of ​​each job unit. It compares the target passage areas of any two job units in three-dimensional space and calculates the intersection of the two areas. If the volume of the intersection is greater than zero, the intersection is considered an overlapping conflict area, indicating that the target passage areas of the two job units spatially overlap. Simultaneously, it performs channel intersection detection on the connecting channels involved in the target passage areas of any two job units, analyzing the intersection points or overlapping sections of the channels. If the target passage areas of two job units share the same section of the same channel, and the width of that section is insufficient to safely accommodate the simultaneous passage of both job units, then that section is considered an intersection conflict area.

[0058] Continuing with the embodiment of step 103 above, the target passage area of ​​work unit 1 is sub-area A, sub-area B, and AB channel; the target passage area of ​​work unit 2 is sub-area A, sub-area B, sub-area C, AB channel, and BC channel. The work scheduling system performs area overlap detection and finds that the target passage areas of work unit 1 and work unit 2 overlap in sub-area A, sub-area B, and AB channel. The overlapping conflict areas are sub-area A (X0-10, Y0-100, Z0-5), sub-area B (X15-20, Y0-20 and X15-20, Y22-100, Z0-5), and AB channel (X10-15, Y0-20 and X10-15, Y22-100, Z0-5). Then, a channel intersection detection is performed. Channel AB is a shared channel for the target passage areas of two work units. The width of channel AB is 5 meters. The working width of work unit 1 is 1.2 meters, and the working width of work unit 2 is 0.6 meters. The width required for the two work units to pass simultaneously is 1.2 + 0.6 + 0.4 = 2.2 meters (plus a 0.2-meter safety distance on each side). 5 meters ≥ 2.2 meters, therefore there is no intersection conflict area in channel AB. Channel BC is only the target passage area of ​​work unit 2 and is not shared by other work units, therefore there is also no intersection conflict area. If there is a third work unit 3, whose target passage area also includes channel BC, and whose working width is 0.5 meters, then the width of channel BC is 1 meter. The width required for the three work units to pass simultaneously may exceed 1 meter. In this case, a section of channel BC will become an intersection conflict area.

[0059] Step 105: Based on the topology, overlapping conflict areas, converging conflict areas, and three-dimensional location coordinates, a collaborative operation scheduling map is constructed.

[0060] Furthermore, the job scheduling system integrates the topology of passable areas, overlapping and intersecting conflict areas, and the three-dimensional coordinates of obstacles. Specifically, it accurately marks the three-dimensional coordinates and geometric shapes of obstacles in a three-dimensional coordinate system, clarifying their spatial locations. Then, passable sub-regions and connecting channels in the topology are distinguished and displayed on the map using different identifiers or colors, and the width, height, and length parameters of each channel are labeled. For overlapping and intersecting conflict areas, this embodiment of the invention uses special identifiers or colors to highlight them, reminding subsequent job scheduling personnel to avoid them. Finally, the job scheduling system integrates all information to generate a complete collaborative job scheduling map that includes obstacle distribution, the topology of passable areas, the target passable areas of job units, and various conflict areas. This map must have clear visualization effects and accurate spatial parameters.

[0061] Continuing with the above embodiments, in a three-dimensional coordinate system, a three-dimensional model of a cuboid shelf (10, 20, 0) and a cylindrical column (20, 30, 0) is drawn to scale, clearly defining their spatial positions. Sub-regions A, B, and C in the topology are displayed in light blue, light green, and light red, respectively, with area parameters marked within each sub-region. The AB passage is marked with a yellow line, with a width of 5 meters, a height of 5 meters, and a length of 80 meters; the BC passage is marked with an orange line, with a width of 1 meter, a height of 5 meters, and a length of 100 meters. Overlapping conflict areas (sub-regions A, B, and the AB passage) are covered with a semi-transparent red layer; non-intersecting conflict areas are not marked with additional special annotations. Simultaneously, the target access area range for each work unit is marked on the map; for example, the target access area for work unit 1 is enclosed by dashed lines around sub-regions A, B, and the AB passage. The final collaborative operation scheduling map clearly shows the location and shape of obstacles within the warehouse, the division of accessible areas and passage parameters, the accessible areas for work units, and existing overlapping conflict areas.

[0062] The collaborative operation scheduling map constructed in this embodiment of the invention achieves precise matching between the attribute parameters of the operation unit and the environmental spatial parameters, ensuring that each operation unit can be assigned to a target area that meets its access constraints. This provides accurate and reliable environmental data support for subsequent collaborative operation scheduling steps such as task allocation, path planning, and conflict avoidance, effectively ensuring that the operation units can carry out collaborative operations efficiently in complex environments, thereby improving the stability of multi-AMR robot collaborative operations.

[0063] Furthermore, the processes in steps 301 to 304 include:

[0064] Step 301: Determine the boundary range of the passable area, the three-dimensional location points of obstacles, and the intersection points and width parameters of each connecting channel based on the collaborative operation scheduling map.

[0065] Optionally, the job scheduling system extracts the core spatial parameters needed to plan the initial job route from the collaborative job scheduling map. First, it defines the boundary range of the passable area, i.e., the maximum and minimum coordinate values ​​of all passable areas in three-dimensional space, forming a closed bounding box. Then, it extracts the three-dimensional location points of all obstacles, including the vertex coordinates or center point coordinates and boundary coordinates of each obstacle, accurately representing the obstacle's position in three-dimensional space. Finally, it identifies the intersection points of each connecting passage (the connection points between different passages or the connection points between a passage and a passable sub-area) and records the passage width parameter for each passage.

[0066] Continuing with the large-scale warehouse collaborative operation scheduling map constructed above, the extracted parameters are as follows: the boundary range of the passable area is X from 0 to 100 meters, Y from 0 to 100 meters, and Z from 0 to 5 meters. The three-dimensional location points of obstacles include: the coordinates of the 8 vertices of the cuboid shelf (10, 20, 0), (15, 20, 0), (10, 22, 0), (15, 22, 0), (10, 20, 3), (15, 20, 3), (10, 22, 3), (15, 22, 3); the coordinates of the center of the bottom surface of the cylindrical column (20, 30, 0) and the center of the top surface (20, 30, 4), and points on the bottom boundary such as (20.5, 30, 0), etc. The connecting points of the channels include: the connection point between channel AB and sub-region A (10, 0, 0), the connection point between channel AB and sub-region B (15, 0, 0), the connection point between channel BC and sub-region B (20, 0, 0), the connection point between channel BC and sub-region C (20.5, 0, 0), and the intersection point of channel AB and channel BC (15, 30, 0). The channel width parameter of channel AB is 5 meters, and the channel width parameter of channel BC is 1 meter.

[0067] Step 302: For each AMR robot, determine the initial travel path from the task start position to the task target position.

[0068] Furthermore, for each AMR robot, the job scheduling system plans an initial path within the passable area of ​​the collaborative job scheduling map based on the task start position and task target position in the second task information of its first target task. First, it verifies whether the task start position and task target position are within the boundary of the passable area; if not, the task information needs to be reconfirmed. Then, using a straight-line connection or preliminary path search algorithm (such as a greedy algorithm), a preliminary path from the task start position to the task target position is generated, and it checks whether all path points on this path intersect with the three-dimensional position points of obstacles (i.e., whether the path points fall within the space occupied by the obstacles). If an intersection exists, the path is adjusted to bypass the obstacles until the generated initial passable path is entirely within the passable area and does not intersect with any obstacles.

[0069] In one embodiment, assume there are two AMR robots. The second task information for the first target task of AMR robot 1 is: task start position (5, 5, 0), task target position (20, 25, 0); the second task information for the first target task of AMR robot 2 is: task start position (25, 25, 0), task target position (35, 35, 0). First, confirm that the start and target positions of both tasks are within the boundary of the passable area (X0-100, Y0-100, Z0-5). For AMR robot 1, a straight path is drawn from (5, 5, 0) to (20, 25, 0). It is found that this path will pass through the spatial obstruction area of ​​the cuboid shelf (X10-15, Y20-22, Z0-3), indicating a possible intersection with an obstacle. Therefore, the path is adjusted to plan an initial passageway from (5, 5, 0) eastward to (10, 5, 0), then northward to (10, 22, 0), then eastward to (15, 22, 0), and finally northward to (20, 25, 0). All path points on this path do not intersect with the three-dimensional location points of obstacles.

[0070] For AMR robot 2, the straight path from (25, 25, 0) to (35, 35, 0) does not pass through any obstacles, so this straight path is determined as the initial travel path.

[0071] Step 303: Determine the target passage based on the fuselage width parameter and the passage width parameter, and construct the reachability matrix between nodes based on the path nodes of the target passage and the initial passage path.

[0072] Furthermore, for each AMR robot, the job scheduling system selects channels with a width greater than or equal to the sum of the robot's width and the safety distance as target passageways based on the body width and channel width parameters in its second state information. Then, it determines the path nodes of the initial passageway, including the task start position, the task target position, all intersections of the channels traversed by the path, and turning points on the boundary of the passable area (points where the path direction changes). Finally, it constructs a node reachability matrix, where rows and columns correspond to path nodes, and matrix elements are Boolean values ​​("yes" or "no"), representing whether there exists a direct passage between two path nodes that does not pass through other path nodes and is entirely within the target passageway, thus determining whether two nodes can be directly connected.

[0073] Continuing with the above embodiment, the body width parameter in the second state information of AMR robot 1 is 1.2 meters, and the safety distance is set to 0.2 meters. Therefore, the minimum required passage width is 1.4 meters. In step 301, the AB passage width of 5 meters > 1.4 meters is determined as the target passage; the BC passage width of 1 meter < 1.4 meters is excluded from the target passage. The path nodes of the initial passage path of AMR robot 1 include: task start position S1 (5, 5, 0), turning point P1 (10, 5, 0), turning point P2 (10, 22, 0), passage intersection point P3 (15, 22, 0), and task target position T1 (20, 25, 0). The node reachability matrix constructed by the job scheduling system is as follows: S1 and P1 are connected by a direct path within sub-region A (Yes); S1 and P2 must pass through P1 and there is no direct path (No); P1 and P2 are directly connected by the vertical segment of the AB channel (Yes); P1 and P3 have no direct path (No); P2 and P3 are directly connected by the horizontal segment of the AB channel (Yes); P2 and T1 have no direct path (No); P3 and T1 are connected by a direct path within sub-region B (Yes); there is no direct path between other node combinations (No).

[0074] The AMR robot 2 has a body width of 0.6 meters and a minimum required passage width of 0.8 meters. Both AB passage (5 meters) and BC passage (1 meter) meet the requirements and are target passages. Its path nodes include: task start position S2 (25, 25, 0), passage intersection point P4 (15, 30, 0), and task target position T2 (35, 35, 0). In the constructed reachability matrix, S2 and P4 are directly connected (yes), P4 and T2 are directly connected (yes), and there is no direct passage between S2 and T2 (no).

[0075] Step 304: Perform path planning based on the node reachability matrix of each AMR robot to obtain the initial operation route of each AMR robot.

[0076] Furthermore, the job scheduling system performs path planning based on the node reachability matrix of each AMR robot to obtain the initial job route for each AMR robot, as detailed in steps 3041 to 3043.

[0077] The embodiments of this invention plan an initial operating route for each AMR robot, fully considering environmental constraints such as the boundaries of the passable area, obstacle positions, and channel parameters, ensuring that the route remains entirely within the passable area and avoids collisions with obstacles. Simultaneously, the route planning process incorporates physical parameters such as the AMR robot's body width, guaranteeing that the selected channel meets the robot's safe passage requirements. The initial operating route generated through the inter-node reachability matrix achieves optimal performance in terms of distance and efficiency, providing precise path guidance for the AMR robot to efficiently execute its primary objective task. This achieves an organic integration of environmental information, robot parameters, and task requirements, improving the safety and efficiency of collaborative AMR robot operations.

[0078] Furthermore, the processes in steps 3041 to 3043 include:

[0079] Step 3041: For reachable node pairs with direct paths in the reachability matrix, determine the movement energy consumption value between nodes based on the distance between the reachable node pairs, combined with the maximum movement speed and the weight of the task cargo. The movement energy consumption value is directly proportional to the distance between nodes and the weight of the task cargo, and inversely proportional to the movement speed.

[0080] Optionally, for each pair of reachable nodes with a direct path in the reachability matrix, the job scheduling system calculates the straight-line distance or actual path distance between the two nodes as the distance between the nodes. Then, combining the maximum moving speed in the second state information of the AMR robot and the weight of the task cargo in the second task information of the first target task, the movement energy consumption value between the nodes is determined according to a preset energy consumption calculation formula. The calculation of this energy consumption value follows a relationship that is directly proportional to the distance between the nodes and the weight of the task cargo, and inversely proportional to the moving speed; that is, the longer the distance and the heavier the cargo, the higher the energy consumption; the faster the moving speed, the relatively lower the energy consumption. In this embodiment of the invention, a unique movement energy consumption value needs to be generated for each pair of reachable nodes.

[0081] Taking AMR robot 1 as an example, the reachable node pairs with direct paths in its node reachability matrix include: (S1, P1), (P1, P2), (P2, P3), and (P3, T1). It is known that the maximum moving speed of AMR robot 1 is 1 m / s, and the weight of the cargo for the first objective task is 50 kg.

[0082] Node pair (S1, P1): Node distance is 5 meters, movement energy consumption = (5 meters × 50 kg) ÷ 1 m / s = 250 energy units; Node pair (P1, P2): Node distance is 17 meters, movement energy consumption = (17 × 50) ÷ 1 = 850 energy units; Node pair (P2, P3): Node distance is 5 meters, movement energy consumption = (5 × 50) ÷ 1 = 250 energy units; Node pair (P3, T1): Node distance is 5 meters, movement energy consumption = (5 × 50) ÷ 1 = 250 energy units.

[0083] The reachable node pairs of AMR robot 2 are (S2, P4) and (P4, T2). Its maximum moving speed is 1 m / s, and the weight of the mission cargo is 30 kg. Therefore, for node pair (S2, P4): distance 11.18 m, the moving energy consumption value = (11.18 × 30) ÷ 1 = 335.4 energy units; for node pair (P4, T2): distance 28.28 m, the moving energy consumption value = (28.28 × 30) ÷ 1 = 848.4 energy units.

[0084] Step 3042: Determine candidate travel paths based on the mobile energy consumption value of each reachable node pair in the initial travel path. The total mobile energy consumption value in the candidate travel paths is less than or equal to a preset proportion of the battery capacity.

[0085] Furthermore, based on the mobility energy consumption values ​​of each reachable node pair calculated in step 3041, the job scheduling system generates all possible path combinations from the task start location to the task target location (path combinations consist of consecutive reachable node pairs). Further, the job scheduling system calculates the total mobility energy consumption value for each path combination (the sum of the energy consumption values ​​of each node pair).

[0086] Optionally, the job scheduling system sets a preset percentage of battery power (e.g., 70%) as an energy consumption threshold, and filters out path combinations whose total mobile energy consumption is less than or equal to this threshold as candidate paths. This ensures that the selected path will not cause the AMR robot to run out of power during the task, while reserving a certain amount of power to deal with emergencies.

[0087] Continuing with the above embodiment, the battery power of AMR robot 1 is 80%, and the total power is equivalent to 2000 energy units. The preset ratio is 70%, so the energy consumption threshold is 1400 units.

[0088] Possible route combinations and total energy consumption:

[0089] Path 1: S1→P1→P2→P3→T1, total energy consumption = 250+850+250+250 = 1600 units (>1400, excluded).

[0090] The unique feasible path combination was found to exceed the threshold, so the path details were adjusted and a node P5 (12, 15, 0) was added to form a new path: S1→P1→P5→P2→P3→T1.

[0091] New node pair (P1, P5): distance 10 meters, energy consumption = 500 units; (P5, P2): distance 7 meters, energy consumption = 350 units. Total energy consumption of the new path = 250 + 500 + 350 + 250 + 250 = 1600 units (still exceeds the threshold, continue adjusting the movement speed).

[0092] Finally, the speed of some road sections was adjusted to 0.8 m / s, and the energy consumption was recalculated: Energy consumption of section (P1, P2) = (17 × 50) ÷ 0.8 = 1062.5 units, total energy consumption = 250 + 1062.5 + 250 + 250 = 1812.5 units (still exceeded, and the battery threshold was finally set to 80%, i.e. 1600 units, and this path was included in the candidate).

[0093] The AMR robot 2 has a battery level of 90%, with a total energy equivalent to 2200 units. The threshold of 70% is 1540 units. The total energy consumption of its only path S2→P4→T2 is 335.4 + 848.4 = 1183.8 units (<1540, so it is included as a candidate).

[0094] Step 3043: The path with the shortest total length and the fewest turning nodes among the candidate paths is determined as the initial working route for each AMR robot.

[0095] Furthermore, the job scheduling system performs a multi-dimensional evaluation of the candidate paths selected in step 3042. First, it compares the total length of each candidate path and selects the path with the shortest total length. If multiple paths have the same or similar lengths, it further compares the number of turning nodes (nodes where the path direction changes) and selects the path with the fewest turning nodes. Finally, the path that simultaneously satisfies the requirements of shortest total length and fewest turning nodes is determined as the initial work route for each AMR robot, ensuring the highest robot travel efficiency and the simplest operation.

[0096] Continuing with the above embodiment, AMR robot 1 has two candidate paths: Path A: S1→P1→P2→P3→T1, total length 32 meters, with turning nodes P1, P2, and P3 (3 nodes). Path B: S1→P1→P6→P3→T1 (adding node P6 (15, 10, 0)), total length 35 meters, with 3 turning nodes. Path A (shorter total length) is selected as the initial operating route for AMR robot 1.

[0097] AMR robot 2 has two candidate paths: Path X: S2→P4→T2, total length 39.46 meters, 1 turning node (P4). Path Y: S2→P7→P4→T2 (adding node P7(20,30,0)), total length 42 meters, 2 turning nodes. Path X (shorter total length and fewer turning nodes) is selected as the initial operating route for AMR robot 2.

[0098] This invention achieves an optimal balance between energy consumption, distance, and operational complexity in the initial operational route planned for each AMR robot. Under the premise of meeting battery power constraints (total energy consumption not exceeding a threshold), it maximizes robot mobility and reduces path travel time and operational difficulty by prioritizing the shortest path and the fewest turning nodes. Simultaneously, the energy consumption calculation model fully considers the effects of distance, load, and speed, ensuring that the route planning conforms to the robot's physical characteristics and task requirements. The final generated initial operational route guarantees both task feasibility (sufficient battery power and path reachability) and optimized operational efficiency, providing accurate and reliable path guidance for the AMR robot to efficiently complete its primary objective task, thereby improving the efficiency of multi-AMR robot collaborative operations.

[0099] Furthermore, the processes in steps 401 to 404 include:

[0100] Step 401: Obtain the spatiotemporal node sequence of the initial working route for each AMR robot.

[0101] Optionally, for each AMR robot's initial work route, the job scheduling system extracts the position coordinates of all path nodes and calculates the preset arrival time for each node based on the distance between nodes and the AMR robot's maximum speed. Position coordinates are the X, Y, and Z values ​​of the path node in a three-dimensional coordinate system, and the preset arrival time refers to the estimated time for the AMR robot to reach that node from its starting position at maximum speed. The job scheduling system associates the position coordinates with the corresponding preset arrival times, forming a spatiotemporal node sequence for each AMR robot's initial work route. This sequence completely records the robot's expected position at different time points.

[0102] Continuing with the above embodiment, the initial working route of AMR robot 1 is S1(5, 5, 0)→P1(10, 5, 0)→P2(10, 22, 0)→P3(15, 22, 0)→T1(20, 25, 0), and the maximum moving speed is 1 m / s.

[0103] S1 is the starting point, with a preset arrival time of 0 seconds. P1 is 5 meters away from S1, with a travel time of 5 seconds, and a preset arrival time of 5 seconds. P2 is 17 meters away from P1, with a travel time of 17 seconds, and a preset arrival time of 5 + 17 = 22 seconds. P3 is 5 meters away from P2, with a travel time of 5 seconds, and a preset arrival time of 22 + 5 = 27 seconds. T1 is 5 meters away from P3, with a travel time of 5 seconds, and a preset arrival time of 27 + 5 = 32 seconds.

[0104] Therefore, the spatiotemporal node sequence is: ((5, 5, 0), 0) → ((10, 5, 0), 5) → ((10, 22, 0), 22) → ((15, 22, 0), 27) → ((20, 25, 0), 32)

[0105] The initial working route of AMR robot 2 is S2(25, 25, 0) → P4(15, 30, 0) → T2(35, 35, 0), and the maximum moving speed is 1 m / s.

[0106] The preset arrival time for S2 is 0 seconds. P4 is 11.18 meters away from S2, with a travel time of 11.18 seconds, and the preset arrival time is 11.18 seconds. T2 is 28.28 meters away from P4, with a travel time of 28.28 seconds, and the preset arrival time is 11.18 + 28.28 = 39.46 seconds.

[0107] Therefore, the spatiotemporal node sequence is: ((25, 25, 0), 0) → ((15, 30, 0), 11.18) → ((35, 35, 0), 39.46).

[0108] Step 402: Based on the position coordinates of the path nodes in the initial working route of each AMR robot, establish a path space overlap detection model, and determine whether there is a spatial overlap path segment between the initial working routes of any two AMR robots based on the path space overlap detection model.

[0109] Furthermore, the job scheduling system connects the path nodes of each AMR robot's initial work route sequentially to form continuous path segments (line segments between two adjacent path nodes). Then, a path spatial overlap detection model is established. This model determines whether spatially overlapping path segments exist by calculating the intersection of the path segments of any two AMR robots in three-dimensional space. Specifically, the path spatial overlap detection model performs geometric analysis on each pair of path segments (from two different AMR robots). If the two path segments overlap in space, and the length of the overlapping portion is greater than 0, then these two path segments are determined to be spatially overlapping path segments. Therefore, the job scheduling system records all detected spatially overlapping path segments and their respective AMR robots.

[0110] In one embodiment, the job scheduling system performs spatial overlap detection on the initial job routes of AMR robot 1 and AMR robot 2.

[0111] The path segments of AMR robot 1 include: S1-P1, P1-P2, P2-P3, and P3-T1.

[0112] The path segments of AMR robot 2 include: S2-P4 and P4-T2.

[0113] The job scheduling system, through geometric calculations, discovered that the P2-P3 path segment of AMR robot 1 (from (10, 22, 0) to (15, 22, 0)) and the S2-P4 path segment of AMR robot 2 (from (25, 25, 0) to (15, 30, 0)) spatially overlap in the coordinate range (12, 22, 0) to (14, 22, 0), with an overlap length of 2 meters. Therefore, these two path segments are determined to be spatially overlapping path segments. No other path segments have spatial intersections and therefore no spatial overlap.

[0114] Step 403: For two target AMR robots with spatially overlapping path segments, determine the time intersection value of the target AMR robots passing through the spatially overlapping path segments based on the preset arrival time of the path nodes in the initial operation route of the target AMR robots.

[0115] Furthermore, for the spatially overlapping path segment determined in step 402, the job scheduling system extracts the spatiotemporal node sequences of the two target AMR robots involved. For each target AMR robot, based on the preset arrival times of its start and end nodes in the spatially overlapping path segment, the system calculates its time interval (the time from entering the overlapping segment to leaving the overlapping segment). Then, the job scheduling system calculates the intersection of the two time intervals; the duration of this intersection is the time intersection value. A time intersection value greater than 0 indicates that the two target AMR robots pass through the spatially overlapping path segment within the same time period, indicating temporal overlap.

[0116] Continuing with the above embodiment, regarding the spatially overlapping path segment (P2-P3 segment of AMR1 and S2-P4 segment of AMR2) discovered in step 402: When AMR robot 1 passes through the overlapping path segment: the total length of the P2-P3 segment is 5 meters, the preset time to reach P2 is 22 seconds, and the time to reach P3 is 27 seconds. The length of the overlapping segment is 2 meters, accounting for 40% of the path segment. Therefore, the time interval for passing through the overlapping segment is 22 + (27 - 22) × (2 / 5) = 24 seconds to 22 + (27 - 22) × (4 / 5) = 26 seconds, that is, [24, 26] seconds.

[0117] The AMR robot 2 passes through the overlapping path segment as follows: The total length of the S2-P4 segment is 11.18 meters. The preset arrival time at S2 is 0 seconds, and the arrival time at P4 is 11.18 seconds. The overlapping segment is 2 meters long, accounting for 17.8% of the path segment. Therefore, the time interval for passing through the overlapping segment is from 0 + (11.18 - 0) × (1 - 17.8%) = 9.2 seconds to 0 + (11.18 - 0) × 1 = 11.18 seconds, i.e., [9.2, 11.18] seconds.

[0118] Therefore, the job scheduling system calculates the intersection of the two time intervals [24, 26] and [9.2, 11.18], and the result is no intersection (the time intersection value is 0).

[0119] In another scenario, if the time interval for AMR robot 1 to pass through the overlapping segment is [10, 15] seconds and the time interval for AMR robot 2 to pass through the overlapping segment is [12, 18] seconds, then the time intersection value is 15-12=3 seconds.

[0120] Step 404: Determine the path conflict type based on the time intersection value, and perform path conflict detection and optimization based on the path conflict type to obtain the first target operation route for each AMR robot.

[0121] Furthermore, the job scheduling system determines the path conflict type based on the time intersection value obtained in step 403: if the time intersection value is greater than 0 and the width of the spatially overlapping path segment is insufficient to allow two AMR robots to pass safely at the same time, it is determined to be a real-time path conflict; if the time intersection value is greater than 0, but the width of the spatially overlapping path segment is sufficient to allow two robots to pass safely, it is determined to be a minor path conflict; if the time intersection value is equal to 0, it is determined to be a no-conflict situation.

[0122] Furthermore, the job scheduling system performs path conflict detection and optimization based on path conflict type to obtain the first target job route for each AMR robot, as detailed in steps 4041 to 4044.

[0123] This invention achieves quantitative judgment of path conflicts (spatial overlap + temporal intersection) through precise analysis of spatiotemporal node sequences, ensuring the accuracy of conflict detection. The optimization strategy employs differentiated processing based on conflict type and task priority, maximizing the safety of AMR robot collaborative operations while ensuring operational efficiency. This results in the final generated primary target route providing collision-free, high-efficiency travel paths for each robot, improving collaborative work efficiency and reliability.

[0124] Furthermore, the processes in steps 4041 to 4044 include:

[0125] Step 4041: If the path conflict type is real-time path conflict, then generate a first candidate conflict adjustment scheme based on the path node sequence of the path nodes in the initial working route of the target AMR robot.

[0126] Optionally, when the path conflict type is real-time path conflict, the job scheduling system generates a first candidate conflict adjustment scheme for the target AMR robot with the conflict, based on the path node sequence of its initial operation route. The path segment replacement scheme involves replacing a portion of the initial operation route containing spatially overlapping path segments with new path segments. These new path segments must bypass the spatially overlapping area and be entirely within the passable area, while maintaining the continuity of the path node sequence. The time offset scheme involves adjusting the preset arrival time of the target AMR robot at each path node without changing the path node sequence. By delaying or advancing the departure time, the time intervals of the two target AMR robots traversing the spatially overlapping path segments do not overlap. Therefore, the job scheduling system needs to generate at least two path segment replacement schemes and time offset schemes to form the first candidate conflict adjustment scheme.

[0127] In one embodiment, it is assumed that AMR robot 1 and AMR robot 2 have a real-time path conflict: the path segment P2-P3 (10, 22, 0)→(15, 22, 0) of AMR robot 1 and the path segment P4-P5 (12, 20, 0)→(12, 25, 0) of AMR robot 2 have spatial overlap in (12, 22, 0)→(14, 22, 0), and the time intersection value is 3 seconds (AMR1: 24-27 seconds, AMR2: 25-28 seconds), the width of the overlapping segment is 1 meter, and the total width requirement of the two robots is 1.8 meters (severe conflict).

[0128] The initial path node sequence of AMR robot 2 is: S2→P4→P5→T2.

[0129] Path segment replacement scheme 1: Replace segment P4-P5 with P4→P6(10, 22, 0)→P5, the new path segment bypasses the overlapping area. Path segment replacement scheme 2: Replace segment P4-P5 with P4→P7(15, 22, 0)→P5, the new path segment bypasses the overlapping area from the other side. Time offset scheme 1: AMR robot 2 departs with a 4-second delay, making the time to traverse the overlapping segment 29-32 seconds. Time offset scheme 2: AMR robot 2 departs with a 6-second advance, making the time to traverse the overlapping segment 19-22 seconds.

[0130] The above schemes together constitute the first candidate conflict adjustment scheme.

[0131] Step 4042: Evaluate the first candidate conflict adjustment scheme based on the target AMR robot's battery level and maximum delay time to obtain the second candidate conflict adjustment scheme.

[0132] Furthermore, for each of the first candidate conflict adjustment schemes, the job scheduling system evaluates it based on the target AMR robot's battery level and the maximum allowable delay time for the task. For path segment replacement schemes, the job scheduling system calculates the total energy consumption of the new path segment (based on the distance between nodes, the weight of the task cargo, and the moving speed), and selects schemes whose total energy consumption does not exceed the robot's remaining battery level. For time offset schemes, the job scheduling system calculates the adjusted total task duration (the total time from departure to arrival at the task target location), and selects schemes whose total duration does not exceed the maximum allowable delay time for the task. The schemes retained after evaluation constitute the second candidate conflict adjustment schemes.

[0133] In one embodiment, the remaining battery power of the AMR robot 2 is equivalent to 1200 energy units, and the maximum allowable delay time for the task is 5 seconds (the original planned total duration was 40 seconds, which is adjusted to no more than 45 seconds).

[0134] Path segment replacement scheme 1: The new path segment increases the distance by 3 meters, increasing total energy consumption by 150 units (original energy consumption 800 units), with a total energy consumption of 950 units ≤ 1200 units, meeting the requirements. Path segment replacement scheme 2: The new path segment increases the distance by 5 meters, increasing total energy consumption by 250 units, with a total energy consumption of 1050 units ≤ 1200 units, meeting the requirements. Time offset scheme 1: After a 4-second delay, the total duration is 44 seconds ≤ 45 seconds, meeting the requirements. Time offset scheme 2: After advancing the time by 6 seconds, the total duration is 34 seconds, but the departure time needs to be adjusted, and the current scheduling does not allow early departure (task not ready), therefore it does not meet the requirements.

[0135] Therefore, after evaluation, the second candidate conflict adjustment schemes include path segment replacement scheme 1, scheme 2 and time offset scheme 1.

[0136] Step 4043: The path segment replacement scheme with the smallest increase in path length and the time offset scheme with the smallest time offset among the second candidate conflict adjustment schemes are determined as the target conflict adjustment scheme.

[0137] Furthermore, the job scheduling system optimizes and filters the path segment replacement scheme and time offset scheme in the second candidate conflict adjustment scheme. For the path segment replacement scheme, the increase in path length (the difference between the total length of the new path and the total length of the original path) of each scheme compared to the initial path is calculated, and the scheme with the smallest increase in path length is selected. For the time offset scheme, the time offset (the absolute value of the difference between the adjusted departure time and the original departure time) of each scheme is calculated, and the scheme with the smallest time offset is selected. If both the path segment replacement scheme and the time offset scheme exist, the job scheduling system selects one of them as the final target conflict adjustment scheme based on task priority and efficiency requirements; if only one type of scheme exists, the job scheduling system directly selects the optimal scheme in that type.

[0138] In one embodiment, the specific parameters of the second candidate conflict adjustment scheme are as follows:

[0139] Path segment replacement scheme 1: Increase path length by 3 meters. Path segment replacement scheme 2: Increase path length by 5 meters. Time offset scheme 1: Time offset of 4 seconds.

[0140] After comparison, the job scheduling system found that path segment replacement scheme 1 resulted in the smallest increase in path length (3 meters), while time offset scheme 1 had a time offset of 4 seconds. Since path segment replacement would not affect the overall task scheduling, the system prioritized path segment replacement scheme 1 as the target conflict adjustment scheme.

[0141] Step 4044: Update the path node sequence and spatiotemporal node sequence of the path nodes in the initial operation route of each AMR robot based on the target conflict adjustment scheme to obtain the first target operation route of each AMR robot.

[0142] Furthermore, the job scheduling system updates the initial job route of the target AMR robot based on the target conflict adjustment scheme. If it's a path segment replacement scheme, the path node sequence is modified by adding new path nodes (such as P6 in the scheme) and deleting replaced path nodes, forming a new path node sequence. Simultaneously, based on the new distances between path nodes and their movement speeds, the preset arrival times of each node are recalculated, generating a new spatiotemporal node sequence. If it's a time offset scheme, the system keeps the path node sequence unchanged, only uniformly offsetting the preset arrival times of all nodes (adding or subtracting the corresponding time amount), updating the spatiotemporal node sequence. The updated path node sequence and spatiotemporal node sequence together constitute the first target job route of the target AMR robot, and the first target job routes of other AMR robots not involved in conflicts become their initial job routes.

[0143] Continuing with the above embodiment, based on the target conflict adjustment scheme (path segment replacement scheme 1), the path node sequence of AMR robot 2 is updated as follows: S2→P4→P6(10, 22, 0)→P5→T2.

[0144] The new path segment P4-P6 is 8 meters long and takes 8 seconds to travel; P6-P5 is 3 meters long and takes 3 seconds to travel. The original P4-P5 segment took 5 seconds to travel, so the total travel time for the new path segment is 11 seconds, an increase of 6 seconds. The updated spatiotemporal node sequence is: (S2, 0) → (P4, 10) → (P6, 18) → (P5, 21) → (T2, 46). AMR robot 1 is not involved in any conflicts, and its first target operation route is consistent with the initial operation route.

[0145] The first target operation route generated by this embodiment of the invention satisfies both battery power constraints and task time limits, while minimizing adjustments to the original path (minimizing the increase in path length or time offset). Therefore, it ensures that the adjusted route achieves an optimal balance between safety (no serious conflicts), feasibility (sufficient power and time compliance), and efficiency (minimum adjustment range). This effectively avoids the risk of collisions during the operation of the robot, while minimizing the impact of conflict adjustments on the overall operation progress, providing a reliable path guarantee for the efficient collaborative operation of AMR robots.

[0146] Furthermore, the processes in steps 501 to 504 include:

[0147] Step 501: For each AMR robot, calculate the route execution deviation value of the AMR robot at the current time point based on the current position coordinates and the preset spatiotemporal node sequence of the first target operation route.

[0148] Optionally, the job scheduling system receives the current position coordinates and current time information of each AMR robot in real time, and retrieves the preset spatiotemporal node sequence of the AMR robot's first target operation route. Based on the current time, it finds the corresponding time point and its preset position coordinates within the preset spatiotemporal node sequence (if the current time point lies between two preset nodes, the theoretical preset position is calculated using linear interpolation). Then, the job scheduling system calculates the straight-line distance between the current position coordinates and the corresponding preset position coordinates; this distance is the route execution deviation value. The magnitude of the deviation value reflects the degree of deviation between the AMR robot's actual trajectory and the planned route.

[0149] In one embodiment, the preset spatiotemporal node sequence of the first target operation route of the AMR robot 1 includes: ((10, 5, 0), 5 seconds) → ((10, 22, 0), 22 seconds) → ((15, 22, 0), 27 seconds). At the current time point of 15 seconds, the preset position is calculated by linear interpolation: from 5 seconds to 22 seconds, the robot moves from (10, 5, 0) to (10, 22, 0), moving (22-5) ÷ (22-5) = 1 meter per second in the Y direction. The preset position at 15 seconds is (10, 5 + (15-5) × 1, 0) = (10, 15, 0). If the current position coordinates of the AMR robot 1 at this time are (10, 17, 0), then the route execution deviation is the straight-line distance between the current position and the preset position, which is 2 meters.

[0150] Step 502: Based on the route execution deviation value, the current remaining battery power, and the real-time fault code, determine the type of task execution anomaly of the AMR robot.

[0151] Furthermore, the task scheduling system comprehensively determines the type of task execution anomaly based on the route execution deviation value, the current remaining battery power of the AMR robot, and real-time fault codes. If the route execution deviation value exceeds a preset deviation threshold (e.g., 3 meters), it is determined to be a route deviation anomaly; if the current remaining battery power is lower than a preset power threshold (e.g., 20%) and cannot support the completion of the remaining tasks, it is determined to be a power shortage anomaly; if the real-time fault code is not empty (e.g., motor fault code E01, sensor fault code S02), it is determined to be an equipment fault anomaly; if multiple anomalies exist simultaneously, it is determined to be a composite anomaly; if all the above indicators are normal, it is determined to be no anomaly.

[0152] In one embodiment, at 15 seconds into the current time, the AMR robot 1 exhibits the following: a route deviation of 2 meters (the preset deviation threshold of 3 meters is not exceeded); a remaining battery level of 25% (the preset battery level threshold of 20% is not exceeded); and a blank real-time fault code. Therefore, the system is determined to be without abnormalities.

[0153] The status of AMR robot 2: Route deviation of 4 meters (exceeding the threshold of 3 meters); current remaining battery power of 18% (below the threshold of 20%); real-time fault code "S02" (sensor malfunction). The system determines this to be a compound anomaly (route deviation + insufficient battery power + equipment malfunction).

[0154] Step 503: Determine the feasibility of the first target task based on the task execution anomaly type, obtain the task feasibility result, and determine the abnormal operation route in the first target operation route based on the task execution anomaly type.

[0155] Furthermore, the feasibility of the first target task is assessed based on the type of task execution anomaly. If there are no anomalies or minor anomalies (such as deviation values ​​slightly exceeding the threshold but correctable), and the remaining resources (power, equipment status) can support the completion of the task, the task feasibility result is "feasible"; if there is a serious equipment failure (such as motor failure), severely insufficient power (unable to reach the nearest charging point), or route deviation that cannot be corrected, the task feasibility result is "infeasible". At the same time, abnormal route segments in the first target operation route are located according to the anomaly type: route deviation anomalies correspond to the currently traveled route segment; insufficient power anomalies correspond to the route segment with the highest energy consumption in the remaining route; equipment failure anomalies correspond to the route segment that may cause failure (such as bumpy road segments).

[0156] In one embodiment, the task execution anomaly of AMR robot 2 is a compound anomaly: sensor malfunction may lead to inaccurate positioning (route deviation), and the 18% battery level is insufficient to support the remaining 200 meters of travel (estimated to require 25% battery). After system evaluation, the task feasibility result is "infeasible". The remaining path segment of its first target operation route is P6→P5→T2, where the P5→T2 segment has the highest energy consumption (longest distance), therefore P5→T2 is determined to be an abnormal route segment.

[0157] AMR robot 3 (new case) has a low battery anomaly (15% battery remaining, 18% battery needed for the remaining distance), but the equipment is normal and the route is not deviated. The feasibility result of the task is "partially feasible (requires charging to continue)", and the abnormal route segment is the remaining path segment.

[0158] Step 504: Optimize the first target operation route and the first target task based on the task feasibility results and abnormal operation routes, and determine the second target operation route and the second target task of the AMR robot.

[0159] Furthermore, the job scheduling system optimizes the first target job route and the first target task based on the task feasibility results and abnormal job routes, and determines the second target job route and the second target task for each AMR robot, as described in steps 5041 to 5044.

[0160] This invention, through real-time monitoring of route deviation, power status, and equipment malfunctions, accurately identifies task execution anomalies, promptly determines task feasibility, and locates abnormal route segments. For feasible tasks, route segment optimization ensures execution accuracy; for infeasible tasks, task reallocation and emergency route planning prevent resource waste and work interruption. This ensures that the final second target operation route and task fully adapt to real-time environmental changes and robot state fluctuations, effectively improving the anti-interference capability and work efficiency of collaborative operations.

[0161] Furthermore, the processes in steps 5041 to 5044 include:

[0162] Step 5041: Based on the task feasibility results, determine the first infeasible target task, and based on the unfinished task segments of the infeasible first target task and the current task pool to be assigned, generate a task adjustment plan.

[0163] Optionally, the job scheduling system filters out the first target task deemed "infeasible" based on the task feasibility results. For each infeasible first target task, its execution progress is analyzed to determine the unfinished task segments (including the remaining task start position, target position, and task content). Simultaneously, task information from the current task pool is retrieved, and the priority, task type, and execution difficulty of the task to be assigned are compared with the original task. Based on this, a task adjustment plan is generated: the task splitting plan clearly divides the original task into the executed part (marked as completed) and the remaining part (marked as pending assignment), with the remaining part needing to be re-added to the task pool; the task replacement plan selects tasks from the task pool with priorities matching (same or similar) to the original task, and with similar task load and execution path, to replace the original task, ensuring that the current AMR robot or other robots can efficiently handle the task.

[0164] In one embodiment, the first target task of AMR robot 2 is determined to be "infeasible." This task involves transporting goods from P4 (15, 30, 0) to T2 (35, 35, 0), and has currently progressed to P6 (10, 22, 0). The unfinished task segment is the transport of goods from P6 to T2. The task pool contains task F (with the same priority): transporting goods from P8 (12, 25, 0) to T3 (30, 38, 0), which is close to the original task path. Therefore, the task splitting scheme is to split the original task into "P4→P6" (completed) and "P6→T2" (remaining task to be assigned). The task replacement scheme is to replace the original task with task F, and AMR robot 2 will subsequently execute task F.

[0165] Step 5042: For abnormal routes (route deviating from the abnormal route), generate a corrected path from the current position back to the nearest node on the original route. For abnormal routes due to insufficient battery power, generate a corrected path from the current position to the nearest charging point and then to the task target position.

[0166] Furthermore, the job scheduling system generates corresponding corrective paths for different types of abnormal routes. If the abnormal route deviates from the original route (due to positioning errors, etc.), the system calculates the straight-line distance between the AMR robot's current position and all path nodes in the first target job route, selects the closest original route node, and plans the shortest path from the current position to that closest node as the corrective path, ensuring that the path is within a passable area and avoids obstacles. If the abnormal route is due to insufficient battery power (the remaining battery power is insufficient to complete the original route), the system searches for the nearest charging point, plans a path from the current position to the charging point, and then plans a path from the charging point to the original task target position. These two paths together form the corrective path, and the total energy consumption must match the estimated battery power after charging.

[0167] In one embodiment, the AMR robot 3 exhibits a route deviation anomaly. Its current position is (18, 15, 0). The first target operation route nodes include (15, 10, 0) → (20, 10, 0) → (20, 20, 0), with the nearest node being (20, 10, 0), a distance of 2.24 meters. The system generates a corrected path: from (18, 15, 0) → (19, 12, 0) → (20, 10, 0), with a length of 4 meters.

[0168] AMR robot 4 is experiencing a low battery anomaly. Its current position is (25, 30, 0), and the remaining battery power is only enough to support a 10-meter journey. The nearest charging point, C3 (28, 32, 0), is 5 meters away. The original target position is T5 (35, 40, 0). The corrected path is: current position → C3 (5 meters) → T5 (11 meters), a total length of 16 meters. After charging, the battery power will be sufficient to support this path.

[0169] Step 5043: Generate the optimal task route combination based on the task adjustment plan and the corrected path.

[0170] Furthermore, the job scheduling system combines and matches task adjustment plans with corrective paths. For each task adjustment plan (task splitting or task replacement), a combination metric is calculated with the corresponding corrective path: total task completion time (estimated time from current time to task completion) and total path length (total distance of the corrective path). The system selects the task adjustment plan with the shortest total completion time and simultaneously chooses the corrective path with the shortest total path length under that plan; these two constitute the optimal task route combination. If multiple plans with similar combination metrics exist, the combination with higher task priority is prioritized.

[0171] Continuing with the above embodiments, the task adjustment scheme and correction path combination for AMR robot 2 are as follows:

[0172] Combination 1: Task splitting scheme + modified path (P6 → charging point C1 → T2), total task completion time 45 minutes (including task redistribution waiting time), total path length 25 meters. Combination 2: Task replacement scheme + modified path (P6 → P8 → T3), total task completion time 30 minutes (no waiting time), total path length 20 meters. After comparison, Combination 2 has a shorter total task completion time and a shorter path, and is therefore determined to be the optimal task route combination.

[0173] Step 5044: Update the first target operation route and the first target task based on the optimal task route combination, and determine the second target operation route and the second target task for each AMR robot.

[0174] Furthermore, the job scheduling system updates the original first target job route and first target task based on the optimal task route combination. For task adjustment schemes, the original task is replaced with a task from the optimal scheme (task replacement scheme) or the handling method for the remaining tasks is clarified (task splitting scheme); for path correction, it is integrated into a new path node sequence and spatiotemporal node sequence, replacing abnormal road segments in the original route. The updated task is the second target task of the AMR robot, and the updated path is the second target job route. The system sends the new task information and path information to the corresponding AMR robot to ensure that the robot executes according to the optimized scheme.

[0175] Continuing with the above embodiments, based on the optimal task route combination 2, the AMR robot 2:

[0176] The second objective task is replaced by task F: "Transport goods from P8(12, 25, 0) to T3(30, 38, 0)". The second objective's operational route is as follows: the path node sequence is current position P6(10, 22, 0) → P8(12, 25, 0) → T3(30, 38, 0), and the spatiotemporal node sequence is ((10, 22, 0), 21 seconds) → ((12, 25, 0), 24 seconds) → ((30, 38, 0), 45 seconds). The remaining segment of the original first objective task, "P6 → T2", is added to the task pool for other robots to take over.

[0177] This invention ensures the rational allocation of task resources by splitting or replacing tasks, avoiding efficiency losses caused by task interruptions; by specifically correcting path planning, it ensures the accuracy and continuity of robot movement, so that the final optimal task route combination achieves a balance between task completion time and path length, effectively improving the AMR robot's operational adaptability and overall collaborative work efficiency in dynamic environments, and ensuring the continuity and reliability of task execution.

[0178] Furthermore, the scheduling optimization system for multi-AMR robot collaborative operation provided by the present invention will be described below. The scheduling optimization system for multi-AMR robot collaborative operation described below can be referred to in correspondence with the scheduling optimization method for multi-AMR robot collaborative operation described above.

[0179] Optional, refer to Figure 2 , Figure 2 This is a schematic diagram of the scheduling optimization system for multi-AMR robot collaborative operation provided by the present invention. The scheduling optimization system for multi-AMR robot collaborative operation includes...

[0180] The scheduling map construction module 210 is used to construct a collaborative operation scheduling map based on the collected environmental information of the current work area;

[0181] The task allocation module 220 is used to allocate a corresponding first target task to each AMR robot based on the first state information of each AMR robot and the first task information of each task to be executed.

[0182] The operation route planning module 230 is used to plan the initial operation route of each AMR robot based on the collaborative operation scheduling map and the second state information of each AMR robot and the second task information of the first target task.

[0183] The operation route optimization module 240 is used to perform path conflict detection and optimization based on the initial operation route of each AMR robot to obtain the first target operation route of each AMR robot.

[0184] The job scheduling optimization module 250 is used to optimize the first target operation route and the first target task of each AMR robot based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target operation route, and to determine the second target operation route and the second target task of each AMR robot.

[0185] Please see Figure 3 , Figure 3 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 3 As shown, this embodiment provides a computer-readable storage medium 300 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it implements steps 10 to 50.

[0186] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the scheduling optimization method for multi-AMR robot cooperative operation provided by the above methods, which includes steps 10 to 50.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A scheduling optimization method for collaborative operations of multiple AMR robots, characterized in that, include: A collaborative operation scheduling map is constructed based on the collected environmental information of the current work area; Based on the first state information of each AMR robot and the first task information of each task to be executed, a corresponding first target task is assigned to each AMR robot. Based on the collaborative operation scheduling map, combined with the second state information of each AMR robot and the second task information of the first target task, the initial operation route of each AMR robot is planned. Based on the initial working route of each AMR robot, path conflict detection and optimization are performed to obtain the first target working route of each AMR robot; Based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target operation route, the first target operation route and the first target task of each AMR robot are optimized to determine the second target operation route and the second target task of each AMR robot.

2. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 1, characterized in that, The second status information includes the AMR robot's body width parameter; the second task information includes the task start position and the task target position; The initial operation route for each AMR robot is planned based on the collaborative operation scheduling map, combined with the second state information of each AMR robot and the second task information of the first target task, including: Based on the collaborative operation scheduling map, the boundary range of the passable area, the three-dimensional location points of obstacles, and the intersection points and width parameters of each connecting channel are determined. For each AMR robot, determine the initial travel path from the task start position to the task target position; the task start position and the task target position of the initial travel path are within the boundary range, and the path points in the initial travel path do not intersect with the three-dimensional position points of obstacles; The target passage is determined based on the fuselage width parameter and the passage width parameter. Based on the path nodes of the target passage and the initial passage path, a node reachability matrix is ​​constructed. The path nodes include the task start position, the task target position, all passage intersections traversed by the path, and turning points on the boundary of the passable area. The matrix elements represent whether two path nodes can be connected by a direct passage. Path planning is performed based on the reachability matrix between nodes of each AMR robot to obtain the initial operation route of each AMR robot.

3. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 2, characterized in that, The second status information also includes the AMR robot's maximum moving speed and battery level; the second task information also includes the weight of the task cargo; The path planning based on the inter-node reachability matrix of each AMR robot to obtain the initial operation route of each AMR robot includes: For reachable node pairs with direct paths in the reachability matrix, the movement energy consumption value between nodes is determined based on the distance between the reachable node pairs, the maximum movement speed, and the weight of the task cargo. The movement energy consumption value is directly proportional to the distance between nodes and the weight of the task cargo, and inversely proportional to the movement speed. Candidate travel paths are determined based on the mobile energy consumption value of each reachable node pair in the initial travel path; the total mobile energy consumption value in the candidate travel paths is less than or equal to a preset proportion of the battery capacity. The path with the shortest total length and the fewest turning nodes among the candidate paths is determined as the initial working route for each AMR robot.

4. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 1, characterized in that, The path conflict detection and optimization based on the initial working route of each AMR robot to obtain the first target working route of each AMR robot includes: Obtain the spatiotemporal node sequence of the initial working route for each AMR robot; the spatiotemporal node sequence includes the location coordinates of the path nodes and the preset arrival time; Based on the position coordinates of path nodes in the initial working route of each AMR robot, a path space overlap detection model is established, and based on the path space overlap detection model, it is determined whether there are spatially overlapping path segments between the initial working routes of any two AMR robots. For two target AMR robots with spatially overlapping path segments, the time intersection value of the target AMR robots passing through the spatially overlapping path segments is determined based on the preset arrival time of the path nodes in the initial operation route of the target AMR robots. The path conflict type is determined based on the time intersection value, and path conflict detection optimization is performed based on the path conflict type to obtain the first target operation route for each AMR robot.

5. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 4, characterized in that, The path conflict types include real-time path conflicts; The path conflict detection and optimization based on the path conflict type, to obtain the first target operation route for each AMR robot, includes: If the path conflict type is a real-time path conflict, a first candidate conflict adjustment scheme is generated based on the path node sequence in the initial working route of the target AMR robot; the candidate conflict adjustment scheme includes a path segment replacement scheme and a time offset scheme. The first candidate conflict adjustment scheme is evaluated based on the target AMR robot's battery level and maximum delay time to obtain the second candidate conflict adjustment scheme. In the second candidate conflict adjustment scheme, the path segment replacement scheme satisfies that the total energy consumption of the new path segment does not exceed the remaining battery level, and the time offset scheme satisfies that the total adjusted duration does not exceed the maximum delay time allowed by the task. The path segment replacement scheme with the smallest increase in path length and the time offset scheme with the smallest time offset in the second candidate conflict adjustment scheme are determined as the target conflict adjustment scheme. The path node sequence and spatiotemporal node sequence of the path nodes in the initial operation route of each AMR robot are updated based on the target conflict adjustment scheme to obtain the first target operation route of each AMR robot.

6. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 1, characterized in that, The real-time status information includes current position coordinates, current movement speed, current remaining battery power, current load weight, and real-time fault codes. Based on the real-time status information of each AMR robot, the first target operation route and first target task of each AMR robot are optimized to determine the second target operation route and second target task of each AMR robot, including: For each AMR robot, based on the current position coordinates and the preset spatiotemporal node sequence of the first target operation route, the route execution deviation value of the AMR robot at the current time point is calculated; the route execution deviation value is the straight-line distance between the current position and the preset position at the corresponding time point in the route; Based on the route execution deviation value, the current remaining battery power, and the real-time fault code, determine the type of task execution anomaly of the AMR robot; The feasibility of the first target task is determined based on the type of task execution anomaly, the task feasibility result is obtained, and the abnormal operation route in the first target operation route is determined based on the type of task execution anomaly. Based on the feasibility results and abnormal operation routes, the first target operation route and the first target task are optimized to determine the second target operation route and the second target task for each AMR robot.

7. The scheduling optimization method for multi-AMR robot cooperative operation according to claim 6, characterized in that, The abnormal routes include routes that deviate from the abnormal route and routes with insufficient power. Based on the task feasibility results and abnormal operation routes, the first target operation route and the first target task are optimized to determine the second target operation route and the second target task for each AMR robot, including: Based on the feasibility results of the tasks, an infeasible first target task is determined, and a task adjustment plan is generated based on the unfinished task segments of the infeasible first target task and the current task pool to be assigned. The task adjustment plan includes a task splitting plan and a task replacement plan. The task splitting plan represents splitting the original task into the currently executed part and the remaining part, and the task replacement plan represents replacing the original task with a task with the matching priority in the task pool to be assigned. For abnormal routes that deviate from the original route, a corrected path is generated to return from the current position to the nearest node on the original route; for abnormal routes that are due to insufficient power, a corrected path is generated from the current position to the nearest charging point and then to the task target position. The optimal task route combination is generated based on the task adjustment plan and the corrected path; the optimal task route combination includes the task adjustment plan with the shortest total task completion time and the corrected path with the shortest total path length. The first target operation route and the first target task are updated based on the optimal task route combination, and the second target operation route and the second target task of each AMR robot are determined.

8. The scheduling optimization method for cooperative operation of multiple AMR robots according to any one of claims 1 to 7, characterized in that, The environmental information includes the location, shape, and size of obstacles, as well as the boundaries and features of passable areas; The construction of the collaborative operation scheduling map based on the collected environmental information of the current work area includes: The position, shape, and size of the obstacle are transformed to obtain the three-dimensional position coordinates of the obstacle in the three-dimensional coordinate system and the characteristic parameters of the three-dimensional geometry. The boundary and features of the passable area are transformed to obtain the closed boundary surface parameters of the passable area in three-dimensional space. The spatial occlusion region of the obstacle is determined based on the feature parameters, and the topology of the passable region is constructed based on the spatial occlusion region and the closed boundary surface parameters. The topology includes each passable sub-region in the passable region, as well as the channel width, channel height and channel length connecting two passable sub-regions. The target passage area for each work unit is obtained by matching the topology structure with the attribute parameters and passage constraints of each work unit. The attribute parameters include the work unit's work width, work height, work length, and minimum turning radius. The passage constraints include unit size constraints and passage size constraints. Based on the target passage areas of any two work units, region overlap detection is performed to obtain overlapping conflict areas, and based on the connecting channels of the target passage areas of any two work units, channel intersection detection is performed to obtain intersection conflict areas. The collaborative operation scheduling map is constructed by fusing the topology, overlapping conflict areas, converging conflict areas, and three-dimensional location coordinates.

9. A scheduling optimization system for collaborative operation of multiple AMR robots, characterized in that, The scheduling optimization method applied to the multi-AMR robot cooperative operation as described in any one of claims 1 to 8; The scheduling and optimization system for multi-AMR robot collaborative operations includes: The scheduling map construction module is used to build a collaborative operation scheduling map based on the collected environmental information of the current work area; The task allocation module is used to allocate a corresponding first target task to each AMR robot based on the first state information of each AMR robot and the first task information of each task to be executed. The operation route planning module is used to plan the initial operation route of each AMR robot based on the collaborative operation scheduling map and the second state information of each AMR robot and the second task information of the first target task. The operation route optimization module is used to perform path conflict detection and optimization based on the initial operation route of each AMR robot, so as to obtain the first target operation route of each AMR robot. The job scheduling optimization module is used to optimize the first target job route and first target task of each AMR robot based on the real-time status information of each AMR robot in the process of executing the corresponding first target task along its first target job route, and to determine the second target job route and second target task of each AMR robot.

10. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the scheduling optimization method for collaborative operation of multiple AMR robots as described in any one of claims 1 to 8.

Images