A multi-machine cooperative warehouse management system and method

Abstract

The application discloses a kind of warehouse management system and method of multi-machine cooperation, the system includes in-store middle station, robot end and central control end.The technical scheme of the embodiment of the application, task is handled and scheduled by in-store middle station, robot state information is obtained by robot end, in response to task allocation instruction by central control end, in combination with warehouse map, task priority, task execution location and robot state multiple dimensions, quickly assign the target robot with the highest matching degree of current allocation task, improve the efficiency and rationality of intelligent warehouse multi-robot task allocation, while the path conflict judgment and path conflict resolution of the task execution planning path of target robot and other robot executing other tasks are carried out by central control end, improve the efficiency and safety of multi-robot collaborative operation.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, and in particular to a multi-machine collaborative warehouse management system and method. Background Technology

[0002] With the accelerated upgrading of the warehousing and logistics industry towards intelligence and automation, automated guided vehicles (AGVs) and autonomous mobile robots have become the core carriers of smart warehouse operations. Multi-robot collaborative operation modes are gradually replacing single-robot independent operations, becoming the mainstream method for improving warehouse picking and handling efficiency. However, modern warehouse operations are characterized by dynamically generated tasks, temporary occupation of aisle resources by people / vehicles / goods, and large spatial constraints in the work area. Multiple robots are prone to path conflicts, blocking, and unreasonable task allocation during task execution, placing extremely high demands on the real-time performance of robot scheduling, the dynamic adaptability of path planning, and the conflict resolution capabilities of multi-robot collaboration.

[0003] Existing solutions typically employ a two-tier architecture consisting of a cloud platform and robot terminals. The cloud platform handles both global scheduling and real-time decision-making, resulting in high latency in command transmission and data processing. Furthermore, existing solutions often utilize static rule matching in the task allocation phase. After receiving warehouse tasks, the central control unit primarily assigns robots based on the distance between the robot and the task location or the task's preset fixed priority. This singular decision-making approach leads to inefficient task allocation. Summary of the Invention

[0004] This invention provides a multi-robot collaborative warehouse management system and method to improve the efficiency and rationality of multi-robot task allocation in smart warehouses.

[0005] In a first aspect, embodiments of the present invention provide a multi-machine collaborative warehouse management system, which includes an in-warehouse platform, a robot terminal, and a central control terminal; wherein... The warehouse platform is used to determine at least two pending tasks and warehouse static map data, determine the currently assigned task and the task priority coefficient of the currently assigned task from the pending tasks according to the task priority of the pending tasks, generate a task allocation instruction based on the currently assigned task, the task priority coefficient of the currently assigned task, the task execution location and the warehouse static map data, and send the task allocation instruction to the central control terminal. The robot terminal is used to determine the robot status information of each candidate robot and send the robot status information to the central control terminal. The central control unit is used to respond to the task allocation instruction, determine the state adaptability of each candidate robot based on the received robot state information, and determine the position adaptability of each candidate robot based on the current position of the candidate robot and the task execution location based on the warehouse static map data. It also determines the target robot to execute the currently allocated task from the candidate robots based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability, and performs path conflict judgment and path conflict resolution based on the task execution planning paths of the target robot and other robots executing other tasks.

[0006] Secondly, embodiments of the present invention also provide a multi-machine collaborative warehouse management method, the method comprising: The warehouse middle platform determines at least two pending tasks and warehouse static map data. Based on the task priority of the pending tasks, the current assigned task and the task priority coefficient of the current assigned task are determined from the pending tasks. Based on the current assigned task, the task priority coefficient of the current assigned task, the task execution location, and the warehouse static map data, a task allocation instruction is generated and sent to the central control terminal. The robot status information of each candidate robot is determined by the robot terminal and then sent to the central control terminal. The central control unit responds to the task allocation command, determines the state adaptability of each candidate robot based on the received robot state information, and determines the position adaptability of each candidate robot based on the current position of the candidate robot and the task execution location based on the warehouse static map data. Based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability, the target robot to execute the currently allocated task is determined from the candidate robots, and path conflict judgment and path conflict resolution are performed based on the task execution planning paths of the target robot and other robots executing other tasks.

[0007] The technical solution of this invention processes and schedules tasks through an in-warehouse platform, obtains robot status information through the robot terminal, and responds to task allocation instructions through the central control terminal. Combining multiple dimensions such as warehouse map, task priority, task execution location, and robot status, it quickly assigns the target robot with the highest matching degree to the currently assigned task, thereby improving the efficiency and rationality of multi-robot task allocation in smart warehouses. At the same time, the central control terminal performs path conflict judgment and path conflict resolution on the task execution planning paths of the target robot and other robots performing other tasks, thereby improving the efficiency and safety of multi-robot collaborative operations.

[0008] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of the structure of a multi-machine collaborative warehouse management system provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart of a multi-machine collaborative warehouse management method provided in Embodiment 2 of the present invention. Detailed Implementation

[0011] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0012] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0013] Example 1 Figure 1 This is a schematic diagram of a multi-machine collaborative warehouse management system provided in Embodiment 1 of the present invention. This embodiment is applicable to multi-machine collaborative warehouse management situations, such as... Figure 1As shown, the multi-machine collaborative warehouse management system 100 includes an in-warehouse platform 110, a robot terminal 120, and a central control terminal 130.

[0014] The warehouse platform 110 is used to determine at least two pending tasks and warehouse static map data, determine the currently assigned task and the task priority coefficient of the currently assigned task from the pending tasks according to the task priority of the pending tasks, generate a task allocation instruction based on the currently assigned task, the task priority coefficient of the currently assigned task, the task execution location and the warehouse static map data, and send the task allocation instruction to the central control terminal.

[0015] The multi-machine collaborative warehouse management system in this embodiment adopts a three-level framework, which includes an in-warehouse platform, a central control terminal, and robot terminals. The in-warehouse platform can communicate with the central control terminal and robot terminals via the warehouse intranet.

[0016] In this embodiment, a pending task refers to a task that has not yet been assigned to a robot for execution, such as a pending task of picking or moving goods at a designated storage location in the warehouse. The task priority coefficient indicates the urgency of the pending task. The task execution location is the designated storage location number for the pending task. The currently assigned task is the task currently awaiting robot assignment, determined by the warehouse management platform based on task priority.

[0017] It should be noted that the warehouse static map data in this embodiment is stored in the warehouse central platform. The warehouse static map data is the fixed and unchanging basic warehouse information that has been surveyed in advance. The warehouse static map data may include information such as the distribution of storage locations, aisle dimensions, and restricted areas. Specifically, it includes the specific location, number, and size of shelves and storage locations, the width, length, and turning radius of robot travel aisles, and areas that robots cannot enter, such as fire lanes, equipment maintenance areas, or personnel work areas.

[0018] The task assignment instruction is a signal generated based on the task priority coefficient corresponding to the currently assigned task, the task execution location, and the static map data of the warehouse.

[0019] In practical applications, the warehouse platform can respond to the warehouse task list, and break it down into multiple tasks to be processed. The warehouse task list can include task type, task execution location, and task urgency, etc. The task type can be picking goods, moving goods, or replenishing goods, etc.

[0020] The in-warehouse platform can also assign a task priority coefficient to each pending task based on the overall task urgency. A mapping relationship between task urgency and task priority coefficient can be pre-set, allowing for the rapid determination of the task priority coefficient based on this mapping. Higher task urgency corresponds to a higher task priority coefficient.

[0021] Furthermore, the warehouse platform determines the currently assigned task from the pending tasks based on their task priorities. In this embodiment, tasks with higher priority can be processed first. The warehouse platform can sort the pending tasks from highest to lowest priority, and then select the task with the highest priority as the currently assigned task.

[0022] Furthermore, after determining the current assigned task, the warehouse platform generates a task allocation instruction based on the current assigned task, the task priority coefficient corresponding to the current assigned task, the task execution location, and the warehouse static map data. The task allocation instruction is then sent to the central control terminal so that the central control terminal can assign a robot to perform the task.

[0023] In this embodiment, the central control terminal is specifically used for storing warehouse static map data, receiving warehouse task lists, splitting tasks, sorting task processing order, and determining the currently assigned task. Based on the currently assigned task, warehouse task priority coefficient, task execution location, and stored warehouse static map data, it generates task allocation instructions and sends the task allocation instructions to the central control terminal to realize global task scheduling processing.

[0024] The robot terminal 120 is used to determine the robot status information of each candidate robot and send the robot status information to the central control terminal.

[0025] In this embodiment, the robot can communicate with the warehouse platform, the central control unit, and each candidate robot through the warehouse intranet. The candidate robot is an autonomous mobile robot used for picking and transporting goods. In actual warehousing and logistics scenarios, there will be multiple autonomous mobile robots used for picking and transporting goods at the same time.

[0026] Robot status information is a set of basic parameters that characterize the robot's current real-time operating status and schedulable capabilities. It serves as the basis for task allocation, robot selection, adaptability calculation, and dynamic scheduling. Robot status information can include information such as the robot's task occupancy, remaining battery level, current travel speed, and the location of the warehouse.

[0027] In this embodiment, information such as the task occupancy status, remaining power level, current driving speed, and warehouse location of the candidate robot can be collected through the device terminal, and the robot status information can be obtained in real time by periodically reporting it.

[0028] In this embodiment, the robot's terminal is used to determine the robot status information of each candidate robot and send the robot status information to the central control terminal for data interaction, realizing the robot's own data collection and real-time transmission. In real-world scenarios, the robot's terminal is also used to accurately execute tasks based on task type and path planning.

[0029] The central control terminal 130 is used to respond to the task allocation instruction, determine the state adaptability of each candidate robot based on the received robot state information, and determine the position adaptability of each candidate robot based on the current position of the candidate robot and the task execution location based on the warehouse static map data. It also determines the target robot to execute the currently allocated task from the candidate robots based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability, and performs path conflict judgment and path conflict resolution based on the task execution planning paths of the target robot and other robots executing other tasks.

[0030] In this embodiment, the central control unit can communicate with the robot via the warehouse intranet and the warehouse platform. State adaptability indicates the candidate robot's ability to undertake tasks, location adaptability indicates the distance adaptability between the candidate robot and the currently assigned task, and the target robot is the task execution robot assigned to the currently assigned task.

[0031] The central control unit responds to task allocation instructions sent by the robot terminal and determines the state adaptability of each candidate robot based on the robot's state information. Specifically, it assesses the adaptability of each candidate robot to handle tasks based on the robot's task occupancy and remaining battery level, thus obtaining the candidate robot's state adaptability.

[0032] Furthermore, based on the warehouse static map data, the positional suitability of each candidate robot is determined according to its current position and the task execution location. Specifically, the positional suitability can be determined based on the Euclidean distance between the candidate robot's current position and the task execution location, or based on the warehouse static map data and the shortest traversable distance between the candidate robot's current position and the task execution location. The smaller the shortest traversable distance between the candidate robot's current position and the task execution location, the higher the positional suitability can be set. This embodiment does not impose specific restrictions on the methods for determining state suitability and positional suitability; they can be flexibly set.

[0033] Furthermore, in this embodiment, the target robot to perform the currently assigned task can be determined from the candidate robots by combining the task priority coefficient, the state adaptability and position adaptability of the candidate robots.

[0034] Specifically, in this embodiment, different weights can be set for the task priority coefficient, state adaptability, and position adaptability, and a weighted sum can be performed to obtain the comprehensive matching degree between the candidate robot and the currently assigned task. The candidate robot with the highest comprehensive matching degree is then used as the target robot for the currently assigned task.

[0035] Furthermore, the central control unit in this embodiment can also be used to determine the task execution planning path of the target robot based on static warehouse map data, and to determine and resolve path conflicts based on the task execution planning paths of the target robot and other robots performing other tasks.

[0036] In this embodiment, the central control unit is used to respond to the task allocation instruction, calculate the state adaptability and position adaptability of each candidate robot according to the task allocation instruction, and assign the robot to the current task by combining multiple dimensions such as the task priority coefficient, the state adaptability and position adaptability of the candidate robots. This increases the dimensions of task allocation decision, improves the rationality of robot and task allocation, and enhances task processing efficiency.

[0037] Optionally, the robot status information includes at least the candidate robot's power coefficient, task occupancy coefficient, and robot position. The central control unit is specifically used for: In response to the task allocation instruction, the state adaptability of each candidate robot is determined based on the power coefficient and task occupancy coefficient of each candidate robot sent by the robot terminal; the state adaptability is positively correlated with the power coefficient and negatively correlated with the task occupancy coefficient; the state adaptability represents the degree of adaptability of the candidate robot to undertake tasks. Based on the warehouse static map data, the task distance from each candidate robot to the task execution location is determined according to the task execution location and robot location, and the location adaptability of each candidate robot to the currently assigned task is determined according to the task distance. Based on the state adaptability, task priority coefficient, and position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined; The candidate robot with the highest task matching degree will be selected as the target robot to perform the currently assigned task.

[0038] In this embodiment, the robot status information includes at least the candidate robot's power coefficient, task occupancy coefficient, and robot position. The power coefficient represents the candidate robot's remaining power level, the task occupancy coefficient represents the candidate robot's task occupancy level, and the robot position represents the candidate robot's real-time location in the warehouse. In practical applications, the robot's power coefficient and task occupancy coefficient can be quantized using 0-1 normalization or hierarchical interval quantization, and dynamically calculated by combining the robot's real-time remaining power and task occupancy status to ensure that the power coefficient and task occupancy coefficient accurately reflect the robot's actual status.

[0039] Furthermore, in response to the task allocation command, the central control unit determines the state adaptability of each candidate robot based on the power coefficient and task occupancy coefficient of each candidate robot sent by the robot terminal. The state adaptability is positively correlated with the power coefficient and negatively correlated with the task occupancy coefficient, and can be expressed by the following formula: ; in, For candidate robots State adaptability, The task occupancy coefficient for candidate robots. This represents the power coefficient of the candidate robot.

[0040] Based on the static map data of the warehouse, the accessible task distance from each candidate robot to the task execution location is determined according to the task execution location and robot position. The mapping relationship between task distance and location adaptability is pre-set, and the location adaptability of each candidate robot to the current task is determined according to the mapping relationship and task distance.

[0041] Furthermore, by combining state adaptability, task priority coefficient, and position adaptability, different weights can be set for task priority coefficient, state adaptability, and position adaptability, and a weighted sum can be performed to obtain the task matching degree between the candidate robot and the currently assigned task. The candidate robot with the highest task matching degree is selected as the target robot to execute the currently assigned task.

[0042] In this embodiment, the central control unit is used for data processing. By comprehensively analyzing multi-dimensional information, it calculates the task matching degree between all candidate robots and the currently assigned task, and selects the candidate robot with the maximum task matching degree as the executor of the task to be processed. This allows the target robot for the assigned task to be quickly determined from all candidate robots in the warehouse, thereby improving the efficiency and rationality of task allocation.

[0043] Optionally, based on the state adaptability, task priority coefficient, and position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined, including: Determine the maximum position fit among the position fits of each candidate robot in the warehouse and the current assigned task; The normalized position fit is determined based on the ratio of the position fit of the candidate robot to the maximum position fit. The negative position fit is determined based on the difference between 1 and the normalized position fit. The task matching degree between each candidate robot and the currently assigned task is determined by multiplying the negative return position fitness, state fitness, and task priority coefficient.

[0044] In this embodiment, the maximum position fit is the maximum value among the position fits of each candidate robot to the task execution location of the currently assigned task, the normalized position fit is the ratio of the position fit of the candidate robot to the maximum position fit, and the negative normalized position fit is the difference between 1 and the normalized position fit.

[0045] In this embodiment, the task matching degree between each candidate robot and the currently assigned task is determined based on the state adaptability, task priority coefficient, and position adaptability, which can be expressed by the following formula: ; in, Indicates candidate robots With the currently assigned task Task matching degree This indicates the task priority coefficient for the currently assigned task. This indicates the suitability of the candidate robot for the current assigned task. The maximum fit between each candidate robot in the warehouse and the currently assigned task.

[0046] In this embodiment, by integrating multiple dimensions, including the robot's adaptability to the task, the task priority, and the distance between the robot and the task, the matching degree between the candidate robot and the currently assigned task is calculated, which increases the dimensions of task allocation decision and can improve the rationality of subsequent task allocation.

[0047] Optionally, the robot end is further configured to acquire multimodal obstacle features and send the multimodal obstacle features to the central control end, wherein the central control end is further configured to: Based on the static map data of the warehouse sent by the warehouse central platform, the warehouse is divided into sections to determine the warehouse grid map; the warehouse grid map is composed of grids of a preset size. The obstacle occupancy probability of each grid cell is determined based on the weighted result of the multimodal obstacle features. The passability status of the candidate robot between any adjacent grid is determined based on the obstacle occupancy probability. The movement cost between adjacent grids in the warehouse grid map is determined based on the passability status. Based on the movement cost, determine the task execution planning path that minimizes the cumulative movement cost of the target robot from its current location to the task execution location currently assigned to the task.

[0048] In this embodiment, the robot end is also used to acquire multimodal obstacle features, and the central control end is also used to determine the task execution planning path of the target robot based on the multimodal obstacle features and warehouse static map data sent by the robot end.

[0049] In this embodiment, the multimodal obstacle features are the obstacle occupancy status under multiple perception modes collected by each candidate robot through its own sensors. In practical applications, each candidate robot can be configured with multiple perception mode sensors, such as lidar sensors, visual camera sensors, and ultrasonic sensors.

[0050] The multimodal obstacle features in this embodiment include at least laser obstacle features, visual obstacle features, and ultrasonic obstacle features at various locations in the warehouse.

[0051] Furthermore, based on the static map data of the warehouse, the entire operational area of ​​the warehouse is uniformly divided to construct a standardized warehouse grid map. This warehouse grid map uses fixed-size square units as the basic grid and completes the full coverage modeling of the warehouse environment in the form of a regular grid, distinguishing between communication areas, obstacle areas and cargo areas.

[0052] Furthermore, the obstacle occupancy probability of each grid is determined based on the weighted results of laser obstacle features, visual obstacle features, and ultrasonic obstacle features in the multimodal obstacle features. Based on the grid obstacle occupancy probability, the passable areas and obstacle areas in the static warehouse map are dynamically updated. This dynamic updating of the static warehouse map through grid obstacle occupancy probabilities reflects the actual warehouse environment when the robot is powered on. For example, if an empty aisle is visible on the static map but a temporary pallet is present on-site, the corresponding grid will be marked as an "obstacle area" in real time.

[0053] Specifically, the obstacle occupancy probability of grid (x, y) is obtained by weighted fusion of multimodal obstacle features, and is expressed by the following formula: ; ; in, Let (x, y) be the obstacle features of the grid in the k-th perception mode. Let be the weight of the obstacle feature in the k-th perception modality, and the sum of the weights of the obstacle features in each perception modality is 1.

[0054] It should be noted that the passability status between adjacent grids is either passable or impassable. In this embodiment, the passability status between any two adjacent grids can be determined based on the obstacle occupancy probability.

[0055] Specifically, a preset obstacle threshold is set in advance. If the obstacle occupancy probability of the current grid is greater than or equal to the preset obstacle threshold, it means that the passable state from the current grid to the adjacent grid is impassable; if the obstacle occupancy probability of the current grid is less than the preset obstacle threshold, it means that the passable state from the current grid to the adjacent grid is passable.

[0056] Furthermore, the movement cost between adjacent grids in the warehouse grid map is determined based on the traversability status. In this embodiment, the movement cost between adjacent grids is defined based on the warehouse grid map. If the traversability status between the current grid and the adjacent grid is traversable, the movement cost between the current grid and the adjacent grid is set to 1; if the traversability status between the current grid and the adjacent grid is not traversable, the movement cost between the current grid and the adjacent grid is set to 1000. If the traversability status between the current grid and the adjacent grid is not traversable, the movement cost can be flexibly set to a larger value to facilitate the effectiveness and brevity of subsequent path planning.

[0057] Furthermore, a single-source shortest path algorithm can be used to find the task execution planning path with the minimum cumulative movement cost from the target robot position to the task execution location currently assigned by the task, based on the warehouse grid map and the movement cost between adjacent grids.

[0058] In practical applications, the target robot will continuously interact with the warehouse environment and update the movement cost between adjacent grids in the warehouse grid map. When the robot detects a new dynamic obstacle on the task execution planning path, it can trigger local path replanning. By calculating the movement cost of going around the task execution location from the current robot position, the local path with the minimum detour movement cost is selected.

[0059] In this embodiment, based on the movement cost between adjacent grids in the warehouse grid map, a global shortest path is generated from the robot's location to the task execution location. When new dynamic obstacles are encountered during the task execution path, local path dynamic replanning is performed in conjunction with the movement cost, ensuring the optimality and real-time performance of the task execution planning path from the target robot to the currently assigned task goods. Compared to static planning and simple obstacle avoidance methods, the path planning in this embodiment can make refined path planning decisions based on the probability of dynamic obstacles such as temporary warehouse storage, forklift obstruction, and personnel passage, rationally selecting the task execution planning path and effectively avoiding problems such as unnecessary detours increasing path length and meaningless waiting leading to operation delays. Furthermore, it can be continuously optimized in conjunction with different warehouse operation scenarios, and the adaptability of the path planning continuously improves with the operation duration, perfectly adapting to the operational characteristics of dynamically generated warehouse tasks and real-time environmental changes.

[0060] Optionally, the robot status information also includes the robot speed, and the central control terminal is further specifically used for: Within a preset time range, determine the first task execution planning path of the first robot and the second task execution planning path of the second robot in the warehouse; the first task execution planning path is the task execution plan of the first robot from its current position to the task execution location corresponding to the current task, and the second task execution planning path is the task execution plan of the second robot from its current position to the task execution location corresponding to the current task; wherein, the first robot and the second robot are currently executing different tasks; Based on the intersection of the first task execution planning path and the second task execution planning path, determine the intersection length and intersection position; Based on the positions of the first robot, the second robot, the intersection point, the speed of the first robot, and the speed of the second robot, determine the time difference between the arrival of the first robot and the second robot at the intersection point; Determine the shortest path length between the execution planning path for the first task and the execution planning path for the second task; The conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference. If the conflict determination coefficient is greater than or equal to 1, then it is determined that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within the preset time range.

[0061] In this embodiment, the robot status information also includes the robot speed, and the central control terminal is also used to determine path conflicts based on the task execution planning paths of the target robot and other robots performing other tasks.

[0062] In this embodiment, the first task execution planning path is the task execution planning path of the first robot from its position to the task execution location corresponding to the currently executed task, and the second task execution planning path is the task execution planning path of the second robot from its position to the task execution location corresponding to the currently executed task. The first robot and the second robot are currently executing different tasks.

[0063] In this embodiment, the intersection path is the path where the first task execution planning path and the second task execution planning path coincide. The intersection length is the length of a single intersection path, and the intersection position is the midpoint of a single intersection path. The intersection time difference is the time difference between the first robot's arrival at the intersection path and the second robot's arrival at the intersection path. The conflict determination coefficient is used to represent the degree of path travel conflict between the first robot's first task execution planning path and the second robot's second task execution planning path. The conflict determination coefficient is used to determine path conflict between the first robot's first task execution planning path and the second robot's second task execution planning path.

[0064] It should be noted that the preset time range is a pre-set time length, which can be set according to the maximum time required for the first robot and the second robot to reach their respective task execution locations, without any restrictions.

[0065] Specifically, the first arrival time of the first robot to the intersection position can be determined based on the current position, speed, and intersection position of the first robot. The second arrival time of the second robot to the intersection position can be determined based on the current position, speed, and intersection position of the second robot. Then, the time difference between the first and second arrival times can be used to obtain the intersection time difference between the first and second robots.

[0066] Furthermore, a conflict determination coefficient is determined based on the intersection length of the intersection paths, the shortest path length between the first and second task execution planning paths, and the intersection time difference. The conflict determination coefficient is used to determine path conflicts between the first robot's first task execution planning path and the second robot's second task execution planning path. If the conflict determination coefficient is greater than or equal to 1, it is determined that there is a conflict between the first robot's first task execution planning path and the second robot's second task execution planning path within a preset time range.

[0067] In this embodiment, the conflict determination coefficient is determined by combining the time difference between the arrival of the two robots at the intersection position by the central control terminal. Based on the conflict determination coefficient, the first task execution planning path of the first robot and the second task execution planning path of the second robot are used to determine the path conflict, so as to ensure the accuracy of the path conflict determination, avoid the error of determining the path conflict by relying solely on the path overlap, and improve the accuracy of the path conflict determination.

[0068] Optionally, a conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference, including: The first conflict coefficient is determined based on the ratio of the intersection length to the shortest path length; The second conflict coefficient is determined based on the ratio of a preset time threshold to the intersection time difference; The conflict determination coefficient is determined by multiplying the first conflict coefficient and the second conflict coefficient.

[0069] In this embodiment, the preset time threshold is a pre-set duration, which can be determined by the total time the robot takes to traverse the intersection paths. The first conflict coefficient is the ratio of the intersection length to the shortest path length, the second conflict coefficient is the ratio of the preset time threshold to the intersection time difference, and the conflict determination coefficient is the product of the first and second conflict coefficients.

[0070] In this embodiment, the conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference, and can be expressed by the following formula: ; in, Indicates the first robot First task execution planning path and second robot The conflict determination coefficient of the second task execution planning path. The length of the intersection. The shortest path length between the execution planning paths for the first and second tasks. The time difference between the arrival of the first robot and the second robot at the intersection point. This is a preset time threshold.

[0071] In this embodiment, the conflict determination coefficient is determined by the intersection length of the intersection paths, the shortest path length in the two task execution planning paths, and the time difference between the arrival of the robots with conflicting paths at the intersection position. This increases the decision-making dimension of conflict determination and improves the accuracy of conflict determination.

[0072] Optionally, the central control terminal is also specifically used for: In response to the determination that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within a preset time range, the first task priority coefficient of the first assigned task executed by the first robot and the second task priority coefficient of the second assigned task executed by the second robot are determined. Based on the first task priority coefficient and the first robot speed, the first right-of-way priority of the first robot is determined, and based on the second task priority coefficient and the second robot speed, the second right-of-way priority of the second robot is determined. The candidate robot corresponding to the minimum value between the first right-of-way priority and the second right-of-way priority is determined as the waiting robot; Send a yield queuing instruction to the waiting robot so that the waiting robot waits for a preset time before reaching the intersection position to resolve path conflicts.

[0073] In this embodiment, the central control terminal is also used to resolve path conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within a preset time range, in response to the determination that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot.

[0074] In this embodiment, the first assigned task is the current assigned task of the first robot, and the second assigned task is the current assigned task of the second robot. The first task priority coefficient is the priority coefficient of the first assigned task, and the second task priority coefficient is the priority coefficient of the second assigned task. The first right-of-way priority is the right-of-way priority of the first robot on the intersection path, and the second right-of-way priority is the right-of-way priority of the second robot on the intersection path.

[0075] When the central control terminal determines that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within a preset time range, it determines the first right-of-way priority of the first robot based on the first task priority coefficient and the speed of the first robot, and determines the second right-of-way priority of the second robot based on the second task priority coefficient and the speed of the second robot.

[0076] In this embodiment, the first path priority of the intersection path is determined based on the first task priority coefficient and the first robot speed, and can be determined by the following formula: ; in, For the first robot In the first priority of the intersection path The priority coefficient of the first task assigned to the first robot. For the first robot The first robot speed, The maximum robot speed between the first and second robots. For the first robot The remaining workload, The maximum amount of remaining tasks between the first robot and the second robot.

[0077] The remaining workload of the robot can be obtained from the robot's terminal and sent to the central control terminal.

[0078] In this embodiment, the first right-of-way priority of the first robot and the second right-of-way priority of the second robot on the intersection path are determined by the above formula. The higher the right-of-way priority, the higher the priority for passing through the intersection path.

[0079] Furthermore, the candidate robot corresponding to the minimum value between the first right-of-way priority and the second right-of-way priority is determined as the waiting robot, and a yield queuing instruction is sent to the waiting robot so that the waiting robot waits for a preset time before reaching the intersection position to resolve path conflicts. The preset time can be determined based on the speed of the robot that does not need to wait and the length of the intersection path.

[0080] In this embodiment, a Nash equilibrium conflict resolution strategy can also be adopted, which positively correlates the revenue function with the right-of-way priority and negatively correlates it with the robot's waiting time. This resolves path conflicts between the target robot and other robots performing other tasks, thus avoiding deadlock caused by both robots competing for the right of way.

[0081] In addition, in the scenario of path conflict when two robots meet in a narrow passage, if the paths of the intersection cannot be crossed, a yielding position is set at the adjacent position of the intersection. The robot with lower right-of-way priority enters the yielding position first, and the robot with higher right-of-way priority passes through first and then queues to pass through the intersection of the narrow passage.

[0082] In this embodiment, the central control unit determines the first right-of-way priority of the first robot and the second right-of-way priority of the second robot on the intersection path. By comparing the priority values, the principle of lower priority yields to higher priority is followed to ensure that one robot can pass while the other waits. In high-frequency path conflict scenarios such as intersections and narrow passages, the simple "first-come, first-served" conflict avoidance rule is avoided, as it affects the overall task execution efficiency of robots and the task processing efficiency of urgent tasks. This embodiment effectively improves task execution efficiency and task processing efficiency by resolving path conflicts through right-of-way priority.

[0083] Furthermore, this system also provides an intelligent robot election strategy to resolve local conflict networks. When multiple robots form a local conflict network, a command robot can be elected via a temporary local area network. The election criteria are the robot's right-of-way priority and communication signal strength. The command robot is responsible for negotiating and mediating local conflicts and sending the negotiation results back to the central control unit, realizing local negotiation and local record-keeping.

[0084] This system also provides occupation and timeout release strategies. For core areas such as intersections and narrow passages, occupation locks are set. When a robot occupies a core area, the lock state is Lock=1, and other robots must wait. If a robot occupies a core area for an abnormal period exceeding the timeout threshold, a timeout release is triggered, and the lock state is set to Lock=0. This avoids operational interruptions caused by abnormal occupation of core areas and ensures the continuity of warehouse robot operations.

[0085] This embodiment integrates road right-of-way priority determination, command robot collaboration, occupancy locks, and timeout release mechanisms to create a conflict resolution system. This overcomes the limitations of existing single-scheduling modes, enabling proactive conflict prediction, efficient resolution, and differentiated processing. For different high-frequency conflict scenarios such as intersections, narrow passages, and congested areas, road right-of-way priority determines the passage order, and a balanced strategy identifies the optimal yielding or straight-through options. Furthermore, in complex multi-robot conflict scenarios, a command robot can be elected to achieve local collaboration, fundamentally preventing issues such as lane-jumping, collisions, and deadlocks. The occupancy lock and timeout release mechanisms further prevent operational interruptions caused by abnormal occupancy in core areas, ensuring a high success rate in conflict resolution.

[0086] This embodiment upgrades the traditional passive handling after a conflict to proactive prediction before a conflict occurs, which greatly improves the safety of multi-robot collaborative operations, while reducing the waiting time caused by conflicts and ensuring the continuity of the overall warehouse operation process.

[0087] Optionally, the central control terminal is also specifically used for: After determining the target robot to perform the current assigned task, determine the adjacent task distance between the task execution location corresponding to the current assigned task and the task execution location corresponding to the next assigned task, as well as the subsequent task distance between the robot position of the target robot and the task execution location corresponding to the next assigned task. If the product of the subsequent task distance and the preset along-the-way distance threshold is greater than or equal to the adjacent task distance, then the next task will be assigned to the target robot.

[0088] In this embodiment, the adjacent task distance is the robot's traversable distance between the task execution location of the currently assigned task and the task execution location corresponding to the next assigned task. The subsequent task distance is the robot's traversable distance between the target robot's position and the task execution location corresponding to the next assigned task. The preset along-path distance threshold is a pre-set along-path distance threshold used to describe the threshold when, after executing the current assigned task, the distance from the robot to the task execution location of the next assigned task is less than the distance from the robot's position before executing the current assigned task to the task execution location of the next assigned task.

[0089] In this embodiment, the central control unit can also be used for merging tasks along the same route. For the next assigned task of the currently assigned task of the target robot, the task along the same route is determined based on the distance between the adjacent tasks of the current assigned task and the next assigned task, and the distance between the target robot and the subsequent tasks of the next assigned task. If the product of the distance of the subsequent tasks and the preset distance threshold along the same route is greater than or equal to the distance between the adjacent tasks, it means that after the robot executes the current assigned task, the distance from the robot to the next task execution location is less than the distance from the robot to the next task execution location before the robot executes the current assigned task. The next assigned task can be directly assigned to the target robot for execution. Merging adjacent tasks to be assigned to the same target robot can improve task processing efficiency.

[0090] In this embodiment, the central control unit is also used to determine the tasks along the route and to merge and assign the tasks along the route to the same target robot, thereby improving task processing efficiency and task allocation efficiency.

[0091] Optionally, the central control terminal is also specifically used for: In response to the task allocation instruction, determine whether the task priority coefficient of the currently allocated task is equal to a preset priority threshold; If the task priority coefficient of the currently assigned task is equal to the preset priority threshold, determine the current task priority coefficient of each candidate robot's current task. Any candidate robot whose current task priority coefficient is less than that of the currently assigned task is identified as the target robot, and the execution order of the currently assigned task is set to precede that of the target robot's current task.

[0092] In this embodiment, the preset priority threshold is the maximum value that the task priority coefficient can take in advance as preset by the system. When an emergency task is processed in the warehouse middle platform, its task priority coefficient takes the maximum value, which is the preset priority threshold.

[0093] In this embodiment, the currently executing task is the task that the candidate robot is currently executing, and the current task priority coefficient is the priority coefficient of the task that the candidate robot is currently executing.

[0094] If the priority coefficient of the currently assigned task is equal to the preset priority threshold, it means that the currently assigned task is an urgent task and the target robot needs to be assigned to execute the urgent task immediately.

[0095] If the currently assigned task is an urgent task, a queue-jumping mechanism is triggered. By determining the current task priority coefficient of each candidate robot in the warehouse, any candidate robot whose current task priority coefficient is less than the task priority coefficient of the currently assigned task is selected. If the current task priority coefficient of a candidate robot is less than the task priority coefficient of the currently assigned task, it means that the candidate robot is executing a non-urgent task. In this case, any robot is selected from the above candidate robots as the target robot to execute the currently assigned urgent task, and the execution order of the currently assigned task is set to be before the current task of the target robot to ensure the processing efficiency of the urgent task.

[0096] In this embodiment, the central control unit determines urgent tasks based on whether the task priority coefficient is equal to a preset priority threshold, and identifies the robot currently performing a non-urgent task based on the task priority coefficient, which is then used as the target robot to perform the currently assigned urgent task, thereby improving the allocation and processing efficiency of urgent tasks.

[0097] This invention discloses a multi-robot collaborative warehouse management system, comprising an in-warehouse platform, robot terminals, and a central control unit. The technical solution of this invention uses the in-warehouse platform to process and schedule tasks, acquires robot status information through the robot terminals, and responds to task allocation instructions through the central control unit. By combining multiple dimensions such as warehouse map, task priority, task execution location, and robot status, the system quickly assigns the target robot with the highest matching degree to the currently assigned task, improving the efficiency and rationality of multi-robot task allocation in the smart warehouse. Simultaneously, the central control unit performs path conflict judgment and resolution on the task execution planning paths of the target robot and other robots performing other tasks, enhancing the efficiency and safety of multi-robot collaborative operations. Specific Implementation Taking the multi-machine collaborative management of power material handling in a warehouse as an example, loading and unloading of power materials using robots involves robot scheduling, robot path planning, and multi-machine conflict resolution. In the actual architecture, a three-level framework is adopted, including an in-warehouse platform, a central control unit, and the robot itself. The core steps are as follows: The robot collects multimodal obstacle features and robot status information, and transmits them to the central control unit in real time. The central control unit completes the fusion of multimodal obstacle features to obtain dynamic environmental data and robot status; The central control unit performs the optimal allocation of tasks and robots based on task matching degree, and supports emergency tasks to be skipped and tasks along the way to be merged. The central control unit generates the task execution planning path for the currently assigned task and the target robot, as well as the local replanning path; The robot executes path instructions, and at the same time, it completes state interaction with other robots through local cooperative communication. The central control unit detects and resolves path conflicts, and can issue yield instructions, wait instructions, or detour instructions.

[0099] Specifically, in the scenario of robotic handling in the storage of electrical materials, the following detailed steps are included: Robot startup and self-test (T0-T1 minutes): The robot is powered on and starts up, activating sensors such as LiDAR, vision camera, positioning module, and RFID reader to acquire multi-modal obstacle features of candidate robots in the warehouse under multi-sensory modalities, and collect robot status information, including robot position, remaining battery power, task occupancy, remaining task load, and robot speed.

[0100] Inter-system communication is established (T1-T2 minutes). The robot sends a "power-on ready" signal, multimodal obstacle features, and robot status information to the central control unit through the intranet in the warehouse. At the same time, a low-latency communication link is established with the central control unit. After receiving the signal, the central control unit sends back a "communication confirmation" command. The central control unit then obtains the robot status information and multimodal obstacle features of each candidate robot.

[0101] Environmental data fusion (T2-T3 minutes): The central control unit generates a warehouse grid map with obstacle occupancy probability by combining multimodal obstacle characteristics with the warehouse static map data issued by the warehouse platform through multimodal data fusion basic formulas. This completes the environmental data calibration after system startup and provides basic data for subsequent operations.

[0102] Furthermore, when a warehouse task is initiated (dynamically triggered starting at time T3), the warehouse platform generates a task list (including task type, task execution location, and task priority). Then, based on the task list, the task is split, a task priority coefficient is assigned to each task, and a task allocation instruction for the currently assigned task is generated and sent to the central control terminal. The central control unit's dynamic task allocation module responds to task allocation instructions, combines locally cached robot status information (robot position, power coefficient, task occupancy coefficient), initiates a multi-dimensional task-robot matching process, calculates robot status adaptability, filters available robots with adaptability exceeding the threshold, and then calculates the Euclidean distance between the available robot and the task execution location coordinates as the location adaptability. Based on the task priority coefficient, status adaptability, and location adaptability, the target robot with the maximum matching degree is selected as the optimal executor for the currently allocated task through the task-robot matching degree formula. The central control unit issues task execution instructions (including task details, route merging list, and emergency queue jumping indicator) to the selected target robot, updates the robot task occupancy coefficient, and sends the allocation results back to the warehouse middle platform for record-keeping.

[0103] Furthermore, low-latency dynamic path planning (within 50ms after task allocation) uses the central control unit to define the movement cost between adjacent grids based on the dynamic obstacle occupancy probability after multimodal fusion. Based on the movement cost, robot position, and task execution location, a global basic path planning strategy is initiated to solve for the reachable path with the minimum cumulative movement cost, thus obtaining the global basic path sequence. The model of the warehouse platform can also be called to optimize the path strategy through the reward function, making the path safe and efficient. The central control unit sends the task execution planning path and speed constraint instructions to the robot.

[0104] Robot operation execution (continuous execution after receiving path instructions): The robot receives path instructions and initiates walking, turning, and stopping actions according to the path sequence. It collects its own position deviation data in real time and dynamically corrects its execution posture to ensure accurate arrival at the task execution location. After arriving at the task execution location, it performs grasping / handling / replenishment operations according to the task type. The RFID reader reads the location information and sends it back to the central control terminal for confirmation. During the operation, the robot establishes a temporary local area network with surrounding robots through local collaborative communication, and exchanges heartbeat data such as position, speed, and task progress in real time to provide inter-robot status support for conflict resolution.

[0105] Furthermore, proactive conflict prediction (conducted in real time throughout the entire operation) uses a conflict determination coefficient formula to predict conflict risks in real time and trigger the path conflict resolution process.

[0106] A dual-mechanism conflict resolution system (responding within 30ms after conflict prediction) is implemented. The central control unit determines right-of-way priorities and, based on these priorities, issues yielding and queuing instructions to lower-priority robots. If a complex conflict network exists with three or more robots, a robot election algorithm is used to elect a commanding robot that leads local conflict negotiation, determining detour / waiting solutions. After conflict resolution, the central control unit synchronously updates the path planning of the relevant robots to ensure operational continuity and records the conflict resolution log.

[0107] Dynamic environment adaptation and adjustment (triggered in real time when the environment changes). If the robot detects dynamic obstacles (temporary stacked items, forklifts blocking the way, personnel passing through), it will upload the obstacle information (type, occupancy probability, duration) to the central control terminal. The central control terminal will update the warehouse obstacle data by fusing obstacle features.

[0108] Once the task is completed (after a single task is executed), the robot sends a "task completed" signal to the central control unit and simultaneously uploads task execution data (execution time, path deviation, and cargo location confirmation information). After receiving the signal, the central control unit updates the robot's task occupancy coefficient, summarizes the task execution data and performance indicators, and uploads them to the warehouse platform to complete the task loop.

[0109] Next, the robot shutdown process (work completed / human command triggered): After receiving the human shutdown command or completing the recharging, the robot will drive to the designated parking position and automatically shut down all the robot's sensors, disconnect the communication link between the robot and the central control unit, complete the final status detection and upload it to the central control unit, cut off the operating power supply of the robot, and retain only the sleep power supply to complete the shutdown process.

[0110] Example 2 Figure 2 This is a flowchart illustrating a multi-machine collaborative warehouse management method according to Embodiment 2 of the present invention. Based on the above embodiments, this embodiment provides a multi-machine collaborative warehouse management method, which is applicable to multi-machine collaborative warehouse management situations. Figure 2 As shown, the method includes: S210. Determine at least two pending tasks and warehouse static map data through the warehouse middle platform. Determine the currently assigned task and its priority coefficient from the pending tasks according to their priority. Generate a task assignment instruction based on the currently assigned task, its priority coefficient, the task execution location, and the warehouse static map data. Send the task assignment instruction to the central control terminal.

[0111] S220. Determine the robot status information of each candidate robot through the robot terminal, and send the robot status information to the central control terminal.

[0112] S230. Responding to the task allocation instruction via the central control terminal, the state adaptability of each candidate robot is determined based on the received robot state information, and the position adaptability of each candidate robot is determined based on the current position of the candidate robot and the task execution location based on the warehouse static map data. The target robot for executing the currently allocated task is determined from the candidate robots based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability. Path conflict judgment and path conflict resolution are performed based on the task execution planning paths of the target robot and other robots executing other tasks.

[0113] Optionally, the robot status information includes at least the battery level coefficient, task occupancy coefficient, and robot position of the candidate robots. The central control unit responds to the task allocation instruction by determining the status adaptability of each candidate robot based on the received robot status information, and by determining the position adaptability of each candidate robot based on the current position and the task execution location using the warehouse static map data. Finally, the target robot for executing the currently allocated task is determined from the candidate robots based on the task priority coefficient of the currently allocated task, the status adaptability, and the position adaptability, including: The central control unit responds to the task allocation command and determines the state adaptability of each candidate robot based on the power coefficient and task occupancy coefficient of each candidate robot sent by the robot terminal. The state adaptability is positively correlated with the power coefficient and negatively correlated with the task occupancy coefficient. The state adaptability represents the degree of adaptability of the candidate robot to undertake tasks. Based on the warehouse static map data, the task distance from each candidate robot to the task execution location is determined according to the task execution location and robot location, and the location adaptability of each candidate robot to the currently assigned task is determined according to the task distance. Based on the state adaptability, task priority coefficient, and position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined; The candidate robot with the highest task matching degree will be selected as the target robot to perform the currently assigned task.

[0114] Optionally, based on the state adaptability, task priority coefficient, and position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined, including: Determine the maximum position fit among the position fits of each candidate robot in the warehouse and the current assigned task; The normalized position fit is determined based on the ratio of the position fit of the candidate robot to the maximum position fit. The negative position fit is determined based on the difference between 1 and the normalized position fit. The task matching degree between each candidate robot and the currently assigned task is determined by multiplying the negative return position fitness, state fitness, and task priority coefficient.

[0115] Optionally, the robot terminal is further configured to acquire multimodal obstacle features and send the multimodal obstacle features to the central control terminal. The method further includes: The central control unit divides the warehouse into sections based on the static map data sent by the warehouse platform, thus determining a warehouse grid map; the warehouse grid map consists of grids of a preset size. The obstacle occupancy probability of each grid cell is determined based on the weighted result of the multimodal obstacle features. The passability status of the candidate robot between any adjacent grid is determined based on the obstacle occupancy probability. The movement cost between adjacent grids in the warehouse grid map is determined based on the passability status. Based on the movement cost, determine the task execution planning path that minimizes the cumulative movement cost of the target robot from its current location to the task execution location currently assigned to the task.

[0116] Optionally, the robot status information also includes robot speed, and path conflict determination is performed by the central control terminal based on the task execution planning paths of the target robot and other robots performing other tasks, including: Within a preset time range, the central control unit determines the first task execution planning path of the first robot and the second task execution planning path of the second robot in the warehouse. The first task execution planning path is the task execution planning path of the first robot from its current position to the task execution location corresponding to the current task, and the second task execution planning path is the task execution planning path of the second robot from its current position to the task execution location corresponding to the current task. The first robot and the second robot are currently executing different tasks. Based on the intersection of the first task execution planning path and the second task execution planning path, determine the intersection length and intersection position; Based on the positions of the first robot, the second robot, the intersection point, the speed of the first robot, and the speed of the second robot, determine the time difference between the arrival of the first robot and the second robot at the intersection point; Determine the shortest path length between the execution planning path for the first task and the execution planning path for the second task; The conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference. If the conflict determination coefficient is greater than or equal to 1, then it is determined that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within the preset time range.

[0117] Optionally, a conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference, including: The first conflict coefficient is determined based on the ratio of the intersection length to the shortest path length; The second conflict coefficient is determined based on the ratio of a preset time threshold to the intersection time difference; The conflict determination coefficient is determined by multiplying the first conflict coefficient and the second conflict coefficient.

[0118] Optionally, path conflict resolution can be performed by the central control unit based on the task execution planning paths of the target robot and other robots performing other tasks, including: The central control unit responds to the determination that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within a preset time range, and determines the first task priority coefficient of the first assigned task executed by the first robot and the second task priority coefficient of the second assigned task executed by the second robot. Based on the first task priority coefficient and the first robot speed, the first right-of-way priority of the first robot is determined, and based on the second task priority coefficient and the second robot speed, the second right-of-way priority of the second robot is determined. The candidate robot corresponding to the minimum value between the first right-of-way priority and the second right-of-way priority is determined as the waiting robot; Send a yield queuing instruction to the waiting robot so that the waiting robot waits for a preset time before reaching the intersection position to resolve path conflicts.

[0119] Optionally, the method further includes: After determining the target robot to execute the currently assigned task, the central control terminal determines the adjacent task distance between the task execution location corresponding to the current assigned task and the task execution location corresponding to the next assigned task, as well as the subsequent task distance between the robot position of the target robot and the task execution location corresponding to the next assigned task. If the product of the subsequent task distance and the preset along-the-way distance threshold is greater than or equal to the adjacent task distance, then the next task will be assigned to the target robot.

[0120] Optionally, the method further includes: The central control unit responds to the task allocation instruction and determines whether the task priority coefficient of the currently allocated task is equal to a preset priority threshold. If the task priority coefficient of the currently assigned task is equal to the preset priority threshold, determine the current task priority coefficient of each candidate robot's current task. Any candidate robot whose current task priority coefficient is less than that of the currently assigned task is identified as the target robot, and the execution order of the currently assigned task is set to precede that of the target robot's current task.

[0121] This invention discloses a multi-machine collaborative warehouse management method. The method includes: determining at least two pending tasks and warehouse static map data through an in-warehouse platform; determining a currently assigned task and its priority coefficient from the pending tasks based on their priority; generating a task allocation instruction based on the currently assigned task, its priority coefficient, the task execution location, and the warehouse static map data; and sending the task allocation instruction to a central control unit; determining the robot status information of each candidate robot through a robot terminal and sending the robot status information to the central control unit; responding to the task allocation instruction at the central control unit; determining the state adaptability of each candidate robot based on the received robot status information; determining the position adaptability of each candidate robot based on the warehouse static map data, its current position, and the task execution location; determining a target robot from the candidate robots to execute the currently assigned task based on the task priority coefficient, the state adaptability, and the position adaptability; and performing path conflict judgment and resolution based on the task execution planning paths of the target robot and other robots executing other tasks. The technical solution of this invention processes and schedules tasks through an in-warehouse platform, obtains robot status information through the robot terminal, and responds to task allocation instructions through the central control terminal. Combining multiple dimensions such as warehouse map, task priority, task execution location, and robot status, it quickly assigns the target robot with the highest matching degree to the currently assigned task, thereby improving the efficiency and rationality of multi-robot task allocation in smart warehouses. At the same time, the central control terminal performs path conflict judgment and path conflict resolution on the task execution planning paths of the target robot and other robots performing other tasks, thereby improving the efficiency and safety of multi-robot collaborative operations.

[0122] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

[0123] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0124] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A multi-machine collaborative warehouse management system, characterized in that, The system includes an in-warehouse platform, a robot terminal, and a central control terminal; among which... The warehouse platform is used to determine at least two pending tasks and warehouse static map data, determine the currently assigned task and the task priority coefficient of the currently assigned task from the pending tasks according to the task priority of the pending tasks, generate a task allocation instruction based on the currently assigned task, the task priority coefficient of the currently assigned task, the task execution location and the warehouse static map data, and send the task allocation instruction to the central control terminal. The robot terminal is used to determine the robot status information of each candidate robot and send the robot status information to the central control terminal. The central control unit is used to respond to the task allocation instruction, determine the state adaptability of each candidate robot based on the received robot state information, and determine the position adaptability of each candidate robot based on the current position of the candidate robot and the task execution location based on the warehouse static map data. It also determines the target robot to execute the currently allocated task from the candidate robots based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability, and performs path conflict judgment and path conflict resolution based on the task execution planning paths of the target robot and other robots executing other tasks.

2. The system according to claim 1, characterized in that, The robot status information includes at least the candidate robot's power coefficient, task occupancy coefficient, and robot position. The central control unit is specifically used for: In response to the task allocation instruction, the state adaptability of each candidate robot is determined based on the power coefficient and task occupancy coefficient of each candidate robot sent by the robot terminal; the state adaptability is positively correlated with the power coefficient and negatively correlated with the task occupancy coefficient; the state adaptability represents the degree of adaptability of the candidate robot to undertake tasks. Based on the warehouse static map data, the task distance from each candidate robot to the task execution location is determined according to the task execution location and the robot's location. The location fit between each candidate robot and the currently assigned task is determined according to the task distance. Based on the state adaptability, the task priority coefficient, and the position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined; The candidate robot with the highest task matching degree will be selected as the target robot to perform the currently assigned task.

3. The system according to claim 2, characterized in that, Based on the state adaptability, the task priority coefficient, and the position adaptability, the task matching degree between each candidate robot and the currently assigned task is determined, including: Determine the maximum position fit among the position fits of each candidate robot in the warehouse and the currently assigned task; The normalized position fit is determined based on the ratio of the position fit of the candidate robot to the maximum position fit. The negative position fit is determined based on the difference between 1 and the normalized position fit. The task matching degree between each candidate robot and the currently assigned task is determined by multiplying the negative return position adaptability, the state adaptability, and the task priority coefficient.

4. The system according to claim 1, characterized in that, The robot end is also used to acquire multimodal obstacle features and send the multimodal obstacle features to the central control end, which is further specifically used for: Based on the static map data of the warehouse sent by the warehouse central platform, the warehouse is divided into sections to determine the warehouse grid map; the warehouse grid map is composed of grids of a preset size. The obstacle occupancy probability of each grid cell is determined based on the weighted result of the multimodal obstacle features. The passability status of the candidate robot between any adjacent grid is determined based on the obstacle occupancy probability. The movement cost between adjacent grids in the warehouse grid map is determined based on the passability status. Based on the movement cost, determine the task execution planning path that minimizes the cumulative movement cost of the target robot from its current location to the task execution location currently assigned to the task.

5. The system according to claim 1, characterized in that, The robot status information also includes robot speed, and the central control unit is specifically used for: Within a preset time range, determine the first task execution planning path of the first robot and the second task execution planning path of the second robot in the warehouse; the first task execution planning path is the task execution planning path of the first robot from its current position to the task execution location corresponding to the current task, and the second task execution planning path is the task execution planning path of the second robot from its current position to the task execution location corresponding to the current task; wherein, the first robot and the second robot are currently executing different tasks; Based on the intersection of the first task execution planning path and the second task execution planning path, determine the intersection length and intersection position; Based on the positions of the first robot, the second robot, the intersection position, the speed of the first robot, and the speed of the second robot, determine the time difference between the arrival of the first robot and the second robot at the intersection position; Determine the shortest path length between the execution planning path for the first task and the execution planning path for the second task; The conflict determination coefficient is determined based on the intersection length, the shortest path length, and the intersection time difference; If the conflict determination coefficient is greater than or equal to 1, then it is determined that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within the preset time range.

6. The system according to claim 5, characterized in that, Based on the intersection length, the shortest path length, and the intersection time difference, a conflict determination coefficient is determined, including: The first conflict coefficient is determined based on the ratio of the intersection length to the shortest path length; The second conflict coefficient is determined based on the ratio of a preset time threshold to the intersection time difference; The conflict determination coefficient is determined by multiplying the first conflict coefficient and the second conflict coefficient.

7. The system according to claim 5, characterized in that, The central control unit is also specifically used for: In response to the determination that there is a conflict between the first task execution planning path of the first robot and the second task execution planning path of the second robot within a preset time range, the first task priority coefficient of the first assigned task executed by the first robot and the second task priority coefficient of the second assigned task executed by the second robot are determined. Based on the first task priority coefficient and the first robot speed, the first right-of-way priority of the first robot is determined, and based on the second task priority coefficient and the second robot speed, the second right-of-way priority of the second robot is determined. The candidate robot corresponding to the minimum value between the first right-of-way priority and the second right-of-way priority is determined as the waiting robot; Send a yield queuing instruction to the waiting robot so that the waiting robot waits for a preset time before reaching the intersection position to resolve path conflicts.

8. The system according to claim 1, characterized in that, The central control unit is also specifically used for: After determining the target robot to perform the current assigned task, determine the adjacent task distance between the task execution location corresponding to the current assigned task and the task execution location corresponding to the next assigned task, as well as the subsequent task distance between the robot position of the target robot and the task execution location corresponding to the next assigned task. If the product of the subsequent task distance and the preset along-the-way distance threshold is greater than or equal to the adjacent task distance, then the next task will be assigned to the target robot.

9. The system according to claim 1, characterized in that, The central control unit is also specifically used for: In response to the task allocation instruction, determine whether the task priority coefficient of the currently allocated task is equal to a preset priority threshold; If the task priority coefficient of the currently assigned task is equal to the preset priority threshold, determine the current task priority coefficient of the task currently being executed by each candidate robot; Any candidate robot whose current task priority coefficient is less than the task priority coefficient of the currently assigned task is identified as the target robot, and the execution order of the currently assigned task is set to precede the current task of the target robot.

10. A multi-machine collaborative warehouse management method, characterized in that, The method includes: The warehouse middle platform determines at least two pending tasks and warehouse static map data. Based on the task priority of the pending tasks, the current assigned task and the task priority coefficient of the current assigned task are determined from the pending tasks. Based on the current assigned task, the task priority coefficient of the current assigned task, the task execution location, and the warehouse static map data, a task allocation instruction is generated and sent to the central control terminal. The robot status information of each candidate robot is determined by the robot terminal and then sent to the central control terminal. The central control unit responds to the task allocation command, determines the state adaptability of each candidate robot based on the received robot state information, and determines the position adaptability of each candidate robot based on the current position of the candidate robot and the task execution location based on the warehouse static map data. Based on the task priority coefficient of the currently allocated task, the state adaptability, and the position adaptability, the target robot to execute the currently allocated task is determined from the candidate robots, and path conflict judgment and path conflict resolution are performed based on the task execution planning paths of the target robot and other robots executing other tasks.

Images