Multi-robot path planning method
By collecting and analyzing robot traffic information through a central manager, accurate traffic flow predictions are generated, which solves the problems of conflict and congestion caused by insufficient communication in multi-robot systems and improves task execution efficiency.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- JUXING TECH SHENZHEN CO LTD
- Filing Date
- 2025-11-11
- Publication Date
- 2026-05-21
Smart Images
Figure CN2025134165_21052026_PF_FP_ABST
Abstract
Description
Multi-robot path planning methods
[0001] Cross-references to related applications
[0002] This application is based on and claims priority to Chinese Patent Application No. 202411614066X, filed on November 12, 2024, the entire contents of which are incorporated herein by reference. Technical Field
[0003] This application relates to the field of robotics technology, and in particular to a multi-robot path planning method. Background Technology
[0004] In the field of smart warehousing, intelligent robots have greatly improved the efficiency of logistics operations within warehouses thanks to their autonomous navigation and task execution capabilities. However, as warehouse size increases and task complexity grows, more and more automated warehouses are beginning to adopt multi-robot systems to collaboratively perform tasks.
[0005] In related technologies, multi-robot path planning research is generally adjusted based on whether there are conflicts between the paths of different robots, to ensure that the planned path points of the robots are not within the collision range at the same time.
[0006] However, in actual warehousing environments, the communication capabilities of robots are difficult to guarantee. In highly dynamic environments, if robots fail to make timely adjustments, conflicts or even blockages may occur, reducing the efficiency of robots in performing tasks. Summary of the Invention
[0007] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a multi-robot path planning method, apparatus, system, and storage medium to reduce robot congestion and improve the efficiency of robot task execution.
[0008] In a first aspect, this application provides a multi-robot path planning method, comprising: acquiring travel information of each robot in a target area; determining the travel cost of each robot based on the travel information; establishing a probabilistic model for predicting traffic flow for each robot using the travel information to obtain a traffic flow prediction result; responding to a robot's request information and feeding back the robot's attribute information and corresponding travel cost and traffic flow prediction result, so that the robot can calculate the future travel cost of passing through different locations in the target area based on the traffic flow prediction result, and plan a travel path from the starting point to the target point based on the future travel cost; wherein, the travel information includes at least one of the robot's identifier, location, direction, and path.
[0009] According to one embodiment of this application, determining the travel cost of each robot based on travel information includes: constructing a graph structure based on the target area, wherein the graph structure includes multiple nodes representing intersections between channels and multiple edges representing channels that robots can travel through; calculating the basic travel cost of each edge based on the length of the edge and the robot's moving speed; calculating the historical travel cost of each edge based on the congestion status of each edge within a preset historical time period; identifying the current congestion level of each edge based on the travel information and calculating the real-time travel cost of each edge; wherein the real-time travel cost is updated according to the update time of the robot edge.
[0010] According to one embodiment of this application, the step of using the traffic information to establish a probability model for predicting traffic flow for each robot to obtain a traffic flow prediction result includes: establishing the probability model based on the probability distribution of normal movement of each robot at different nodes at different times and the probability distribution of collisions between each robot at different nodes at different times; determining the current operating status of each robot at each node based on the position, path, and direction of each robot in the traffic information; the operating status includes the probability of collisions between robots at each node and the probability of normal movement of robots at each node; predicting the probability distribution of each robot at different locations in the target area within a future preset time period based on the operating status and the probability model, thereby obtaining a traffic flow prediction result for the target area.
[0011] Secondly, this application provides a multi-robot path planning method, comprising: receiving robot attribute information and corresponding travel cost and traffic flow prediction results fed back by a central manager, and obtaining a target task; establishing a path plan for the target task based on the target task, the travel cost, and the traffic flow prediction results; obtaining information on other robots within the target area and their corresponding traffic information; estimating future travel costs using the path plan and the traffic information of the other robots; calculating the comprehensive cost of the robot's travel path from the starting point to the target point based on the future travel cost and the travel cost fed back by the central manager; and determining the optimal travel path based on the comprehensive cost.
[0012] According to one embodiment of this application, the method further includes: adding robots that meet preset conditions in the passage path as dynamic obstacles to a dynamic obstacle list; wherein the preset conditions are determined according to the task priority of each robot; the task priority is determined according to the time when each robot acquires the task; and replanning the passage path from the starting point to the target point according to the future passage cost and the dynamic obstacle list.
[0013] According to one embodiment of this application, the method further includes: when traveling to the target point according to the travel path, avoiding a target robot passing through the target point.
[0014] Thirdly, this application provides a multi-robot path planning device, comprising: a first acquisition module for acquiring travel information of each robot in a target area; a first determination module for determining the travel cost of each robot based on the travel information; a first establishment module for establishing a probabilistic model of predicted traffic flow for each robot using the travel information to obtain a traffic flow prediction result; and a feedback module for responding to robot requests by feeding back the robot's attribute information and corresponding travel cost and traffic flow prediction result, so that the robot can calculate the future travel cost of passing through different locations in the target area based on the traffic flow prediction result, and plan a travel path from the starting point to the target point based on the future travel cost; wherein the travel information includes at least one of the robot's identifier, location, direction, and path.
[0015] Fourthly, this application provides a multi-robot path planning device, comprising: a receiving module, configured to receive robot attribute information and corresponding travel costs and traffic flow prediction results fed back by a central manager, and to acquire a target task; a second establishing module, configured to establish a path plan for the target task based on the target task, the travel costs, and the traffic flow prediction results; a second acquiring module, configured to acquire information on other robots within the target area and their corresponding traffic information; an estimation module, configured to estimate future travel costs using the path plan and the traffic information of the other robots; a calculation module, configured to calculate the comprehensive cost of the robot's travel path from the starting point to the target point based on the future travel costs and the travel costs fed back by the central manager; and a second determining module, configured to determine the optimal travel path based on the comprehensive cost.
[0016] Fifthly, this application provides a multi-robot path planning system, comprising: a central manager and multiple robots; the central manager is configured to execute the multi-robot path planning method as described in the first aspect above; the robots are configured to execute the multi-robot path planning method as described in the second aspect above.
[0017] In a sixth aspect, this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-robot path planning method as described in the first or second aspect above.
[0018] In a seventh aspect, this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the multi-robot path planning method as described in the first or second aspect above.
[0019] Eighthly, this application provides a chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the multi-robot path planning method as described in the first or second aspect above.
[0020] Ninthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the multi-robot path planning method as described in the first or second aspect above.
[0021] The above-described one or more technical solutions in the embodiments of this application have at least the following technical effects:
[0022] According to the multi-robot path planning method of this application, the central manager collects the travel information of each robot in the target area, enabling the prediction process to take into account the current traffic conditions and generate more accurate traffic flow prediction results. These prediction results are sent to each robot, which then evaluates the travel cost of different paths based on the prediction results, thereby planning the optimal travel path. Since the central manager can independently complete the traffic flow prediction, the dependence on real-time communication is reduced. Even under conditions of limited or unstable communication, the robot can still perform path planning based on the pre-received prediction results, rather than relying on real-time communication feedback. This allows the robot to avoid path selection errors caused by communication delays or interruptions, reduce congestion caused by path conflicts or repeated planning, and improve the efficiency of robot task execution.
[0023] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description
[0024] The above and / or additional aspects and advantages of this application will become apparent and readily understood in conjunction with the following description of the embodiments in conjunction with the accompanying drawings, wherein:
[0025] Figure 1 is an architecture diagram of the multi-robot path planning system provided in an embodiment of this application;
[0026] Figure 2 is a schematic diagram of a multi-robot path planning system in the field of intelligent warehousing provided in an embodiment of this application;
[0027] Figure 3 is one of the flowcharts of the multi-robot path planning method provided in the embodiments of this application;
[0028] Figure 4 is an example of the interaction between the central manager and the robot provided in an embodiment of this application;
[0029] Figure 5 is a schematic diagram of the target region provided in an embodiment of this application;
[0030] Figure 6 is a second flowchart illustrating the multi-robot path planning method provided in this application embodiment;
[0031] Figure 7 is one of the structural schematic diagrams of the multi-robot path planning device provided in the embodiments of this application;
[0032] Figure 8 is a second structural schematic diagram of the multi-robot path planning device provided in an embodiment of this application;
[0033] Figure 9 is a schematic diagram of the structure of the electronic device provided in an embodiment of this application. Detailed Implementation
[0034] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.
[0035] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.
[0036] Driven by automation and digitalization, robotics technology has developed rapidly, and its applications are becoming increasingly widespread. Particularly in the field of smart warehousing, intelligent robots, with their autonomous navigation and task execution capabilities, have significantly improved the efficiency of logistics operations within warehouses. However, as warehouse scale expands and task complexity increases, relying solely on robots independently planning paths is no longer sufficient to meet practical needs. How to utilize effective strategies and algorithms to enable multiple robots to continuously and stably coordinate autonomously in the same environment to avoid collisions, while simultaneously maximizing the overall system's operational efficiency, has become a hot topic in the field of smart warehousing.
[0037] In related technologies, research on multi-robot path planning mainly focuses on resolving potential conflicts between paths, that is, using algorithms to ensure that robots avoid collisions with each other within the same time interval when planning their paths. These algorithms are typically based on predicting the position and speed of each robot, and through calculation and adjustment, ensure that each robot's waypoint does not intersect with the waypoints of other robots at the same time. However, these studies usually assume that the robots have good communication capabilities, enabling them to exchange their position information and path plans in real time, thereby achieving effective path coordination. But in real-world warehouse environments, the communication capabilities of robots are difficult to guarantee. In highly dynamic environments, if robots fail to make timely adjustments, conflicts or even blockages may occur, reducing the efficiency of robot task execution.
[0038] The multi-robot path planning method, apparatus, system, and storage medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.
[0039] First, we introduce the multi-robot path planning system provided in the embodiments of this application. As shown in Figure 1, the multi-robot path planning system 100 may include a central manager 110 and multiple robots.
[0040] The central manager 110 can be a single server or a server cluster comprising multiple servers. Robots can be used in any field, such as mining exploration, emergency rescue, smart warehousing, and industrial automation. Multiple robots can work collaboratively.
[0041] Taking the application of robots in smart warehousing as an example, as shown in Figure 2, the central manager can provide map services, task allocation, data storage, traffic management, and monitoring services. Robots are typically equipped with sensors, cameras, and navigation systems, enabling them to move freely in the warehouse environment, identify the location of goods, and transport them to designated locations. Robots can plan their routes based on the maps and tasks provided by the central manager. Through information interaction, robots and the central manager collaboratively complete the tasks of handling, picking, storing, and retrieving goods.
[0042] The following describes a multi-robot path planning method provided in the embodiments of this application. The multi-robot path planning method can be applied to a terminal, and can be executed by hardware or software in the terminal.
[0043] The terminal includes, but is not limited to, portable communication devices such as mobile phones or tablets with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads). It should also be understood that, in some embodiments, the terminal may not be a portable communication device, but rather a desktop computer with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads).
[0044] The following embodiments describe a terminal including a display and a touch-sensitive surface. However, it should be understood that the terminal may include one or more other physical user interface devices such as a physical keyboard, mouse, and joystick.
[0045] The multi-robot path planning method provided in this application embodiment can be executed by an electronic device or a functional module or entity in an electronic device that can implement the multi-robot path planning method. The electronic devices mentioned in this application embodiment include, but are not limited to, computers, robots, central managers, servers, etc.
[0046] The following describes the multi-robot path planning method provided in this application embodiment, taking the central manager as the execution subject as an example.
[0047] As shown in Figure 3, the multi-robot path planning method includes steps 310, 320, 330 and 340.
[0048] Step 310: Obtain the travel information of each robot in the target area; Step 320: Determine the travel cost of each robot based on the travel information; Step 330: Utilize the travel information to establish a probabilistic model for predicting traffic flow for each robot, in order to obtain traffic flow prediction results; Step 340: Respond to the robot's request information, and provide the robot with its attribute information, corresponding travel cost, and traffic flow prediction results, so that the robot can calculate the future travel cost of passing through different locations in the target area based on the traffic flow prediction results, and plan the travel path from the starting point to the target point based on the future travel cost; wherein, the travel information includes at least one of the robot's identifier, location, direction, and path.
[0049] In this embodiment, the target area can be the area where the robot works. The robot can move freely within the target area to complete various tasks. During its movement within the target area, the robot will report its travel information to the central manager at regular intervals. This travel information can include various types of information about the robot during its movement, such as the robot's identifier, position, direction, and path. Of course, to provide the central manager with more information about the robot's travel, the travel information can also include the robot's starting point, destination, and the travel time at each location.
[0050] The central manager can determine the travel cost of each robot based on traffic information. This cost can be represented by travel time, for example, calculated based on the robot's speed and path. Furthermore, since robots may encounter congestion or conflicts while traversing the target area, the manager can also determine these issues based on individual robot traffic information and combine this data with speed and path information to calculate the robot's travel cost.
[0051] In this embodiment, the central manager can predict the probability distribution of each robot at different locations in the target area within a preset time period based on traffic information, thereby obtaining the traffic flow prediction result for the target area. The preset time period can be any time, such as 10 seconds or 20 seconds in the future; this embodiment does not limit this.
[0052] Specifically, the central manager can use machine learning to predict the probability distribution of each robot at different locations in the target area within a preset time period. For example, a suitable machine learning model can be selected, such as a time series prediction model, support vector machine, random forest, gradient boosting machine, etc. Historical traffic information is used as training data to train the machine learning model. During the training process, the model parameters are updated through the backpropagation algorithm, thereby obtaining a probabilistic model for predicting traffic flow. By inputting traffic information into the probabilistic model, the traffic flow prediction result output by the probabilistic model can be obtained.
[0053] Of course, the central manager can also predict the probability distribution of each robot at different locations within a target area over a predetermined time period in other ways. For example, a probabilistic model for predicting traffic flow can be established based on the physical laws of robot motion (such as speed and acceleration) and environmental constraints. Statistical methods can be used to analyze historical traffic data, identify traffic patterns and trends, apply probability theory principles to estimate the probability distribution of robots at different locations, abstract the target area into a graph structure, and use graph theory algorithms, such as shortest path algorithms, to predict robot travel paths and simulate robot movement within the target area. Then, based on expert knowledge or empirical rules, robot behavior and traffic flow can be predicted, thus obtaining the traffic flow prediction results.
[0054] In this embodiment, after receiving a request from the robot, the central manager can provide the robot with its attribute information, corresponding travel costs, and traffic flow prediction results. The robot's attribute information may include its robot ID, path information, the time it acquired the task, its starting point, and its destination. The robot can determine its task, starting point, and destination based on this attribute information, and then plan its own travel path based on its location, task, starting point, and destination, combined with the traffic flow prediction results.
[0055] Specifically, the robot can calculate the future travel costs through different locations within a target area based on traffic flow predictions. These future travel costs provide insights into the anticipated traffic conditions for the upcoming time period, helping the robot anticipate and avoid potential congestion. They represent the estimated time the robot will take to traverse different locations within the target area. During the analysis of traffic flow predictions, different future travel costs can be set based on the predicted congestion levels at different locations within the target area, including predicted congestion points and vacant areas. For example, higher future travel costs can be set for locations with higher congestion levels, and lower future travel costs for locations with lower congestion levels. After determining the future travel costs at different locations within the target area, pathfinding algorithms, such as Dijkstra's algorithm, A* algorithm, and Bellman-Ford algorithm, can be used to find the path with the lowest future travel cost from the robot's starting point to the target point, based on the robot's origin and target point locations.
[0056] According to the multi-robot path planning method of this application, the central manager collects the travel information of each robot in the target area, enabling the prediction process to take into account the current traffic conditions and generate more accurate traffic flow prediction results. These prediction results are sent to each robot, which then evaluates the travel cost of different paths based on the prediction results, thereby planning the optimal travel path. Since the central manager can independently complete the traffic flow prediction, the dependence on real-time communication is reduced. Even under conditions of limited or unstable communication, the robot can still perform path planning based on the pre-received prediction results, rather than relying on real-time communication feedback. This allows the robot to avoid path selection errors caused by communication delays or interruptions, reduce congestion caused by path conflicts or repeated planning, and improve the efficiency of robot task execution.
[0057] In some embodiments, determining the passage cost of each robot based on passage information includes: constructing a graph structure based on the target area, wherein the graph structure includes multiple nodes representing intersections between channels and multiple edges representing channels that robots can pass through; calculating the basic travel cost through each edge based on the length of the edge and the robot's moving speed; calculating the historical passage cost of each edge based on the congestion status of each edge within a preset historical time period; identifying the current congestion level of each edge based on the passage information and calculating the real-time passage cost of each edge; wherein the real-time passage cost is updated according to the update time of the robot edge.
[0058] In this embodiment, as shown in Figure 4, the central manager can calculate the basic travel cost, historical travel cost, and real-time travel cost respectively, and send the basic travel cost, historical travel cost, real-time travel cost, and traffic flow prediction results to each robot.
[0059] In some embodiments, as shown in Figure 5, the target area can be abstracted as a graph structure, which includes multiple nodes and multiple edges; wherein, the edges represent passageways that the robot can pass through, and the nodes represent the intersections between passageways, used to describe the robot's position in the target area.
[0060] In this embodiment, the locations of shelves, picking points, packing stations, and aisles can be set according to the map dimensions of the target area. To improve traffic speed and reduce robot congestion, all road segments can be defined as one-way roads, as shown in Figure 5. The target area can be abstracted into a graph structure as follows:
[0061] G = (F, E)
[0062] Where G represents the graph structure, F represents the set of nodes in graph G, and E represents the set of edges in graph G.
[0063] In graph G, each edge can be labeled with a numerical value that has a certain meaning; this value is called the weight of the edge. Such a graph is called a weighted graph. To make the weights more meaningful, this embodiment defines the weight as the estimated time required for the robot to traverse each edge, measured in seconds.
[0064] In this embodiment, by abstracting the target area into a graph structure, the complex spatial environment is simplified into a series of computable nodes and edges, which can clearly identify and quantify the attributes of channels and intersections, thereby facilitating path search and traffic prediction and improving the accuracy and efficiency of path planning.
[0065] In this embodiment, the basic travel cost across each edge can be calculated based on the edge length and the robot's moving speed. For any edge e = (i,j) ∈ E, the travel time is determined by the edge length and the robot's moving speed in the absence of interference from other robots. The basic travel cost of edge e is Cost. b (e) is:
[0066] Where d(i,j) represents the travel distance between nodes i and j, and V represents the robot's moving speed.
[0067] In this embodiment, the central manager calculates the length of the edge and the robot's moving speed to determine the basic travel cost of each edge. This allows the robot to consider not only traffic conditions but also physical distance when planning its path. The robot can then identify the path with the lowest cost, thereby ensuring traffic efficiency while avoiding congested areas and reducing waiting time.
[0068] In this embodiment, in a smart warehousing environment, road congestion is often determined by both the map structure and the number of tasks. These factors are typically stable over a period of time. The warehouse's past congestion levels can provide a reference for robot path planning, allowing it to avoid congested areas as much as possible. The preset historical time period can be the past 20 seconds, the past 50 seconds, or other times. The historical travel cost of each edge can be calculated based on the congestion status of each edge within the preset historical time period; that is, the higher the congestion level, the higher the historical communication cost.
[0069] In some embodiments, it can be based on the formula
[0070] Calculate the historical passage cost for each edge;
[0071] Cost h (e) represents the historical passage cost of edge e, H(e) is the time occupancy rate of edge e during observation time T, represents the congestion state of edge e during observation time T, N represents the number of observations, and t j This represents the time it takes for robot j to traverse edge e each time.
[0072] In this embodiment, the observation time T is a preset historical time period. The congestion status of each edge within the preset historical time period is represented by the time occupancy rate of each edge within the observation time, thereby calculating the historical passage cost of each edge.
[0073] In this embodiment, by incorporating the analysis of historical congestion status into path planning and calculating the historical passage cost of each side, the robot not only considers traffic flow prediction when planning its path, but also incorporates the understanding of congestion patterns based on historical data. This allows the robot to identify areas that have historically been congested and proactively avoid these areas during planning, thereby enabling the robot to better cope with dynamically changing traffic environments and further reduce congestion.
[0074] When robots encounter conflicts, they trigger local obstacle avoidance strategies, which significantly increases the time it takes for them to traverse the conflict point. If the rate of congestion mitigation is slower than the rate of new congestion, the congested area expands, severely impacting the operational efficiency of the robot system. To avoid large-scale robot congestion, real-time travel costs are introduced to reflect road congestion, and these costs are continuously updated using an exponential smoothing method.
[0075] In this embodiment, the passage information may also include the passage time of each robot through each edge. The central manager 110 can determine the current congestion level of each edge based on the passage time of each edge. The longer the passage time, the higher the current congestion level of the edge, which can increase the real-time passage cost. The shorter the passage time, the lower the current congestion level of the edge, which can reduce the real-time passage cost.
[0076] In some embodiments, it can be based on the formula
[0077] Update the real-time passage cost of each edge;
[0078] in, Let λ represent the real-time travel cost of edge e in the nth update, and let t represent the update parameter. n (e) represents the time taken for the robot to pass through edge e for the nth time, indicating the current congestion level of edge e.
[0079] In this embodiment, the current congestion level of each edge is identified by analyzing traffic information in real time, and the real-time passage cost of these edges is updated accordingly, reflecting the actual time cost incurred by the robot to pass through each edge at the current moment. By combining these real-time passage costs with predicted future passage costs, the robot can evaluate and select the lowest-cost passage path based on the latest congestion information, more effectively avoiding current congested areas, reducing waiting time, and improving traffic efficiency.
[0080] In some embodiments, if the preset duration of edge e is not updated, according to the formula
[0081] Update the real-time passage cost of edge e;
[0082] Cost d (0) represents the initial real-time passage cost, α represents the attenuation coefficient, τ represents the attenuation period, and t represents the preset duration.
[0083] In this embodiment, since the real-time passage cost of a certain edge has not been updated for a long time in a dynamically changing traffic environment, the data may no longer accurately reflect the current traffic conditions. By attenuating the real-time passage cost, the impact of the real-time passage cost on path planning is reduced, the path planning errors caused by data update lag are reduced, and the robot is prevented from making decisions that may no longer be applicable based on outdated information, thereby improving the accuracy of path planning.
[0084] In this embodiment, by abstracting the target area into a graph structure, the complex spatial environment is simplified into a series of computable nodes and edges. This allows for clear identification and quantification of the attributes of channels and intersections. By combining basic travel costs, historical travel costs, and real-time travel costs, the travel costs of each robot are determined. This enables the robot to consider not only traffic conditions when planning its path, but also physical distance, historical data on congestion patterns, and real-time travel costs of each edge based on its current congestion level. This allows the robot to identify the path with the lowest cost, thereby ensuring travel efficiency while avoiding congested areas and reducing waiting time.
[0085] In some embodiments, traffic information is used to establish a probabilistic model for predicting traffic flow for each robot to obtain traffic flow prediction results. This includes: establishing a probabilistic model based on the probability distribution of normal movement of each robot at different nodes at different times and the probability distribution of collisions between each robot at different nodes at different times; determining the current operating status of each robot at each node based on the position, path, and direction of each robot in the traffic information; the operating status includes the probability of collisions between robots at each node and the probability of normal movement of robots at each node; and predicting the probability distribution of each robot at different locations in the target area within a future preset time period based on the operating status and the probabilistic model to obtain the traffic flow prediction results for the target area.
[0086] In this embodiment, a probabilistic model representing traffic flow can be established based on a graph structure. The probability of a robot encountering a conflict at each node and the probability of a robot moving normally at each node can be determined based on the position, path, and direction of each robot in the traffic information. The movement trend of each robot can be identified, and the movement of the robot in the target area can be simulated. This allows the prediction of the probability distribution of each robot at different positions in the target area within a preset time period in the future, thus obtaining the traffic flow prediction result for the target area.
[0087] Specifically, it can make Let X represent the probability that robot i is at node n. i (t) represents the probability distribution of robot i at time t across different nodes. The robot's operation can be categorized into the probability distribution of collisions. The probability distribution N of the robot moving normally at each node. i (t). This represents the probability that robot i will encounter a conflict at node n.
[0088] The relationship between the probability distribution of robot i at different nodes and the robot's operating status is expressed as follows:
[0089] X i (t)=N i(t)+C i (t)
[0090] Based on the robot's operational characteristics under normal and conflict states, and utilizing the property of probability flow conservation, the robot's evolution at the next moment can be simulated, yielding a probabilistic model:
[0091] Among them, X i X(t) represents the probability distribution of robot i at different nodes at time t. i (t+1) represents the probability distribution of robot i at different nodes at time t+1, N i (t) represents the probability distribution of robot i moving normally at different nodes at time t, C i (t) represents the probability distribution of robot i encountering a conflict at different nodes at time t. Let be an n-dimensional matrix, representing the probability that robot i excludes waiting for an actual transfer from one node to another. Let I be an n-dimensional matrix, representing the probability that robot i will move from one node to another after resolving a conflict. Let I be an n-dimensional identity matrix, ψ be the conflict resolution coefficient, and n be the number of nodes.
[0092] The probability that robot i will transition from node u to node v after conflict resolution can be calculated as follows:
[0093] Where A(u,w) represents the adjacency matrix, node w is the adjacent node of node u, and S i (u,v) represents the robot's score during the transition from node u to node v, which can be estimated as:
[0094] Among them, T i Let H(v,T) be the target point of robot i. i () represents the distance from node v to target point T i The shortest path length, H min (u,T i The distance from the target point T among all neighboring nodes of node u can be calculated. i The minimum value of the shortest path length.
[0095] ξ i is the confidence coefficient for robot path i, reflecting the probability that the robot will choose the current path to move, and is defined as:
[0096] ξ i =1-γΔt 2
[0097] Where Δt is the time difference between the robot's last successful communication with the central manager 110, and γ is the confidence decay coefficient.
[0098] Based on the above formula, it can be calculated that...
[0099] Secondly, W i It is an n-dimensional diagonal matrix. This represents the probability that robot i will wait in place at node j. To do this, we need to calculate the probability matrix W′ of the robot waiting when it moves to different nodes at time t. i :
[0100] Where ρ is a fixed waiting coefficient, path(t) is the node where the robot is located at time t, and c v (t) represents the probability that the robot will encounter a conflict at node v at time t, where T i Let δ and ε be the target point of robot i, and let δ and ε be adjustable parameters.
[0101] It is an n-dimensional matrix. The probability that robot i excludes waiting for an actual transfer from node u to node v can be calculated as follows:
[0102] Through the above calculation process, the probability distribution X at different nodes at time t+1 can be calculated. i (t+1).
[0103] In this embodiment, after obtaining the traffic flow prediction results for the target area, the traffic flow prediction results can be sent to each robot. Each robot can calculate its future travel costs through each edge based on the traffic flow prediction results.
[0104] Specifically, the probability distributions of each robot at different locations within the target area can be summed to obtain the future traffic flow matrix X. l×n This matrix represents the total probability distribution of the presence of robots at each node within a predetermined time period l. Based on this probability distribution, robots can determine the congestion situation of each edge within the predetermined time period l, and thus calculate the future travel cost of each edge. For example, if nodes u and v are adjacent nodes, and e is an edge between nodes u and v, then the higher the total probability of nodes u and v within the predetermined time period l, the higher the future travel cost of edge e.
[0105] In this embodiment, by considering the probabilities of robot conflict and normal movement, and combining them with a probabilistic model to predict traffic flow, the likelihood of robot conflict at each node is quantified, thereby identifying potential congestion points and traffic bottlenecks. Furthermore, assessing the probability of normal movement helps determine the robot's passage efficiency in conflict-free situations, providing another dimension of information for traffic flow prediction. Combining the probabilities of robot conflict and normal movement enables the probabilistic model to more comprehensively simulate and predict the dynamic changes in traffic flow, thereby improving the accuracy of traffic flow prediction.
[0106] In some embodiments, the operation status of each robot within a preset time period can be updated based on traffic flow prediction results and probability models.
[0107] In this embodiment, the operational status of each robot within a preset time period can be updated based on traffic flow prediction results and a probability model to continuously predict future traffic flow. Specifically, this can be achieved by using the probability distribution X of the robot at different nodes at time t+1. m (t+1) Further calculate the probability that robot i will encounter a conflict at time t+1.
[0108] Where, x i (t+1) represents the probability that robot m is at different nodes at time t+1.
[0109] The distribution of conflict states at time t+1 is determined by the unresolved conflict states at time t and the newly formed conflict states, as follows:
[0110] We can obtain the probability distribution N of robot i moving normally at different nodes at time t+1. i (t+1)=X i (t+1)-C i (t+1).
[0111] By repeating the prediction multiple times, the traffic flow prediction result within the interval (t, t+l) can be obtained, where l is the prediction duration.
[0112] In this embodiment, by combining traffic flow prediction results and a probability model to update the operation of each robot within a preset time period in the future, this dynamic update mechanism significantly enhances the predictive ability of the system. It can ensure that the operation of the robots is synchronized with the actual traffic conditions and reflect the dynamic changes in the target area in a timely manner, so that the probability model can continuously provide traffic flow predictions.
[0113] The following uses a robot as the executor to illustrate the multi-robot path planning method provided in this application.
[0114] As shown in Figure 6, the multi-robot path planning method includes steps 610, 620, 630, 640, 650, and 660.
[0115] Step 610: Receive robot attribute information and corresponding travel cost and traffic flow prediction results from the central manager, and obtain the target task; Step 620: Based on the target task, travel cost, and traffic flow prediction results, establish a path plan for the target task; Step 630: Obtain information on other robots within the target area and their corresponding traffic information; Step 640: Use the path plan and traffic information of other robots to estimate future travel costs; Step 650: Based on the future travel costs and the travel costs fed back by the central manager, calculate the comprehensive cost of the robot's travel path from the starting point to the target point; Step 660: Determine the optimal travel path based on the comprehensive cost.
[0116] In this embodiment of the application, as shown in Figure 4, the robot can perform path planning based on the information obtained from the central manager. For example, the robot can determine the task, starting point, and target point based on attribute information, and then plan its own travel path based on its own location, task, starting point, and target point, combined with traffic flow prediction results.
[0117] Specifically, the robot can calculate the future travel costs through different locations within a target area based on traffic flow predictions. These future travel costs provide insights into the anticipated traffic conditions for the upcoming time period, helping the robot anticipate and avoid potential congestion. They represent the estimated time the robot will take to traverse different locations within the target area. During the analysis of traffic flow predictions, different future travel costs can be set based on the predicted congestion levels at different locations within the target area, including predicted congestion points and vacant areas. For example, higher future travel costs can be set for locations with higher congestion levels, and lower future travel costs for locations with lower congestion levels. After determining the future travel costs at different locations within the target area, pathfinding algorithms, such as Dijkstra's algorithm, A* algorithm, and Bellman-Ford algorithm, can be used to find the path with the lowest future travel cost from the robot's starting point to the target point, based on the robot's origin and target point locations.
[0118] In this embodiment, the robot can calculate the comprehensive cost of its travel path from the starting point to the destination by combining future travel costs, basic travel costs, historical travel costs, and future travel costs. When planning its path, the robot can choose the path with the lowest comprehensive cost from the starting point to the destination. Specifically, weights can be assigned to each cost type, and then the weighted sum of these costs can be calculated to obtain the comprehensive cost.
[0119] cost(e) = w1cost b (e)+w2cost h (e)+w3cost d (e)+w4cost f (e)
[0120] Where cost(e) represents the total cost of edge e. b (e) represents the basic travel cost of edge e. h (e) represents the historical travel cost of edge e. d (e) represents the real-time travel cost of edge e, cost f (e) represents the future travel cost of edge e, and w1, w2, w3, and w4 are the proportional coefficients of each cost.
[0121] The robot optimizes its journey from origin to destination by combining future travel costs, basic travel costs, historical travel costs, and real-time travel costs. This comprehensive approach allows the robot to make more accurate and comprehensive path selections based on multi-dimensional cost data. Future travel costs provide insights into expected traffic conditions in the upcoming timeframe, helping the robot anticipate and avoid potential congestion. Basic travel costs represent travel time under ideal, congestion-free conditions, providing a benchmark for the robot's travel. Historical travel costs reflect past travel experience under similar conditions, increasing the stability and reliability of path planning. Real-time travel costs represent the current traffic situation, allowing the robot to consider the current context. Through this multi-factor fusion approach to path planning, the robot can achieve more efficient, flexible, and adaptive travel strategies, optimizing its performance in complex traffic environments.
[0122] In some embodiments, the probability distributions of each robot at different locations in the target area within a preset time period can be summed based on the traffic flow prediction results to obtain a future traffic flow matrix; the future traffic flow matrix represents the total probability distribution of the presence of robots at each node; and the future travel cost through each edge can be calculated based on the future traffic flow matrix.
[0123] Specifically, it can be done according to the formula
[0124] Calculate the future travel cost through each edge;
[0125] in, X represents the total probability that a robot exists at node j during time interval t1, where j is the target node connected to edge e. j(t) represents the total probability that a robot exists at node j at time t, t0 represents the prediction start time, l represents the preset time period, z represents the time coefficient, and cost is... f (e) represents the future travel cost through edge e. This represents the estimated cost coefficient.
[0126] In this embodiment, a future traffic flow matrix is generated by using traffic flow prediction results. This matrix obtains the total probability distribution of the presence of robots at each node by summing the probability distribution of each robot at different locations in the target area. This allows for a quantitative and probabilistic assessment of future traffic conditions, making the calculation of future travel costs more accurate.
[0127] According to the multi-robot path planning method of this application, the central manager collects the travel information of each robot in the target area, enabling the prediction process to take into account the current traffic conditions and generate more accurate traffic flow prediction results. These prediction results are sent to each robot, which then evaluates the travel cost of different paths based on the prediction results, thereby planning the optimal travel path. Since the central manager can independently complete the traffic flow prediction, the dependence on real-time communication is reduced. Even under conditions of limited or unstable communication, the robot can still perform path planning based on the pre-received prediction results, rather than relying on real-time communication feedback. This allows the robot to avoid path selection errors caused by communication delays or interruptions, reduce congestion caused by path conflicts or repeated planning, and improve the efficiency of robot task execution.
[0128] In some embodiments, the method further includes: adding robots that meet preset conditions in the passage path as dynamic obstacles to a dynamic obstacle list; wherein the preset conditions are determined according to the task priority of each robot; the task priority is determined according to the time when each robot acquires the task; and replanning the passage path from the starting point to the target point based on the future passage cost and the dynamic obstacle list.
[0129] In this embodiment, after the robot has planned a passage path, if it encounters other robots during its journey, and if it determines that the other robots meet the preset conditions, it can treat these robots as dynamic obstacles and use the Astar algorithm with a time dimension or other path search algorithms to replan the passage path, so that the new passage path does not conflict with the obstacles and the passage cost is minimized.
[0130] The preset conditions can be pre-defined conditions. For example, robots with shorter travel paths can be assigned higher priority. If other robots have higher priorities, then the robot that meets the preset conditions is added to the dynamic obstacle list. Alternatively, different robots can have their travel priorities pre-set. If other robots have higher travel priorities, then the robot that meets the preset conditions is added to the dynamic obstacle list. The preset conditions can be determined based on the task priorities of each robot; the task priorities are determined based on the time each robot acquires its task.
[0131] In this embodiment, the preset conditions can be determined by the task priority of each robot. The robot that gets the task earlier has a higher priority. If the task priority of other robots is higher, it is determined that the robot meets the preset conditions and the robot is added to the dynamic obstacle list as a dynamic obstacle.
[0132] For example, preset conditions may include:
[0133] Rule 1: When two robots' paths cross each other's starting points, the lower-priority robot must give way, and the higher-priority robot ignores the lower-priority robot.
[0134] Rule 2: When a high-priority robot's path passes through the starting point of a low-priority robot and the low-priority robot has not passed through the starting point of the high-priority robot, the high-priority robot must give way, and the low-priority robot ignores the high-priority robot.
[0135] Rule 3: When a low-priority robot's path passes through the starting point of a high-priority robot and the high-priority robot has not passed through the starting point of the low-priority robot, the low-priority robot must give way, and the high-priority robot ignores the low-priority robot.
[0136] In this embodiment, robots that meet preset conditions in the passage path are identified as dynamic obstacles. The preset conditions can be determined by the task priority of each robot. The robot that gets the task earlier has a higher priority. If the task priority of other robots is higher, it is determined that the robot meets the preset conditions and is added to the list of dynamic obstacles. Then, the passage path is replanned based on the future passage cost and this list, so that the robot can take into account the current position and expected behavior of other robots, thereby more accurately predicting and avoiding potential conflicts and congestion, further reducing traffic congestion and improving the efficiency of robot task execution.
[0137] In some embodiments, the method further includes: when traveling to the target point according to the travel path, avoiding a target robot passing through the target point. In this embodiment, after the robot travels to the target point, it may need to wait for task execution at the target point. For example, in a smart warehousing application scenario, the robot also needs to wait at the target point for the picker to complete the picking operation. To avoid affecting the passage of other robots, the robot can avoid the target robot passing through the target point when other robots need to pass through the target point. The target robot can be another robot that needs to pass through the target point, or a robot with a specific identifier.
[0138] In this embodiment, after the robot travels to the target point, it may need to wait at the target point to perform the task. In order to avoid affecting the passage of other people nearby, it can avoid target robots that have passed through the target point, thereby improving passage efficiency.
[0139] The multi-robot path planning method provided in this application can be executed by a multi-robot path planning device. This application uses the execution of the multi-robot path planning method by a multi-robot path planning device as an example to illustrate the multi-robot path planning device provided in this application.
[0140] This application also provides a multi-robot path planning device.
[0141] As shown in Figure 7, the multi-robot path planning device includes: a first acquisition module 710, used to acquire the travel information of each robot in the target area; a first determination module 720, used to determine the travel cost of each robot based on the travel information; a first establishment module 730, used to establish a probabilistic model of the predicted traffic flow for each robot using the travel information, so as to obtain the traffic flow prediction result; and a feedback module 740, used to respond to the robot's request information, and feed back the robot's attribute information and the corresponding travel cost and traffic flow prediction result, so that the robot can calculate the future travel cost of passing through different locations in the target area based on the traffic flow prediction result, and plan the travel path from the starting point to the target point based on the future travel cost; wherein, the travel information includes at least one of the robot's identifier, position, direction, and path.
[0142] According to the multi-robot path planning device of this application, the central manager collects the travel information of each robot in the target area, so that the prediction process takes into account the current traffic conditions and generates more accurate traffic flow prediction results. These prediction results are sent to each robot, and the robot then evaluates the travel cost of different paths based on the prediction results, thereby planning the optimal travel path. Since the central manager can independently complete the prediction of traffic flow, the dependence on real-time communication is reduced. Even in the case of limited or unstable communication, the robot can still perform path planning based on the pre-received prediction results, rather than relying on real-time communication feedback. This allows the robot to avoid path selection errors caused by communication delays or interruptions, reduce congestion caused by path conflicts or repeated planning, and improve the efficiency of robot task execution.
[0143] In some embodiments, the first determining module 720 is further configured to: construct a graph structure based on the target area, wherein the graph structure includes multiple nodes representing intersections between channels and multiple edges representing channels that the robot can pass through; calculate the basic travel cost through each edge based on the length of the edge and the robot's moving speed; calculate the historical travel cost of each edge based on the congestion status of each edge within a preset historical time period; identify the current congestion level of each edge based on the passage information, and calculate the real-time travel cost of each edge; wherein the real-time travel cost is updated according to the update time of the robot edge.
[0144] In some embodiments, the first establishing module 730 is further configured to: establish a probability model based on the probability distribution of normal movement of each robot at different nodes at different times and the probability distribution of conflict between each robot at different nodes at different times; determine the current operation status of each robot at each node based on the position, path and direction of each robot in the traffic information; the operation status includes the probability of conflict between robots at each node and the probability of normal movement of robots at each node; predict the probability distribution of each robot at different locations in the target area within a preset time period in the future based on the operation status and the probability model, and obtain the traffic flow prediction result of the target area.
[0145] This application embodiment also provides a multi-robot path planning device. As shown in FIG8, the multi-robot path planning device includes: a receiving module 810, used to receive robot attribute information and corresponding travel costs and traffic flow prediction results fed back by a central manager, and to acquire a target task; a second establishing module 820, used to establish a path plan for the target task based on the target task, travel costs, and traffic flow prediction results; a second acquiring module 830, used to acquire information on other robots in the target area and their corresponding traffic information; an estimation module 850, used to estimate future travel costs using the path plan and traffic information of other robots; a calculation module 840, used to calculate the comprehensive cost of the robot's travel path from the starting point to the target point based on the future travel costs and the travel costs fed back by the central manager; and a second determining module 860, used to determine the optimal travel path based on the comprehensive cost.
[0146] According to the multi-robot path planning device of this application, the central manager collects the travel information of each robot in the target area, so that the prediction process takes into account the current traffic conditions and generates more accurate traffic flow prediction results. These prediction results are sent to each robot, and the robot then evaluates the travel cost of different paths based on the prediction results, thereby planning the optimal travel path. Since the central manager can independently complete the prediction of traffic flow, the dependence on real-time communication is reduced. Even in the case of limited or unstable communication, the robot can still perform path planning based on the pre-received prediction results, rather than relying on real-time communication feedback. This allows the robot to avoid path selection errors caused by communication delays or interruptions, reduce congestion caused by path conflicts or repeated planning, and improve the efficiency of robot task execution.
[0147] In some embodiments, the second determining module 860 is further configured to: add robots that meet preset conditions in the passage path as dynamic obstacles to a dynamic obstacle list; wherein the preset conditions are determined according to the task priority of each robot; the task priority is determined according to the time when each robot acquires the task; and replan the passage path from the starting point to the target point according to the future passage cost and the dynamic obstacle list.
[0148] In some embodiments, the multi-robot path planning device further includes an avoidance module for avoiding a target robot passing through the target point when traveling to the target point according to the travel path.
[0149] The multi-robot path planning device in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television set (TV), ATM, or self-service machine, etc. This application embodiment does not specifically limit the device.
[0150] The multi-robot path planning device in this application embodiment can be a device with an operating system. This operating system can be a Microsoft (Windows) operating system, an Android operating system, an iOS operating system, or other possible operating systems; this application embodiment does not specifically limit it.
[0151] In some embodiments, as shown in FIG9, this application embodiment also provides an electronic device 900, including a processor 901, a memory 902, and a computer program stored in the memory 902 and executable on the processor 901. When the program is executed by the processor 901, it implements the various processes of the above-described multi-robot path planning method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.
[0152] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.
[0153] This application also provides a multi-robot path planning system, including: the aforementioned central manager and multiple aforementioned robots.
[0154] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described multi-robot path planning method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.
[0155] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.
[0156] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described multi-robot path planning method.
[0157] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.
[0158] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described embodiments of the multi-robot path planning method and achieve the same technical effect. To avoid repetition, it will not be described again here.
[0159] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.
[0160] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.
[0161] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.
[0162] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.
[0163] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0164] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.
Claims
1. A multi-robot path planning method, wherein, include: Obtain the movement information of each robot in the target area; Based on the passage information, determine the passage cost for each robot; Using the traffic information, a probabilistic model for predicting traffic flow is established for each robot to obtain traffic flow prediction results; In response to the robot's request, the system feeds back the robot's attribute information and corresponding travel costs and traffic flow prediction results, so that the robot can calculate the future travel costs through different locations in the target area based on the traffic flow prediction results, and plan the travel path from the starting point to the target point based on the future travel costs. The access information includes at least one of the following: the robot's identifier, location, direction, and path.
2. The method of claim 1, wherein, Determining the passage cost of each robot based on the passage information includes: A graph structure is constructed based on the target region, wherein the graph structure includes multiple nodes representing intersections between channels and multiple edges representing channels that the robot can pass through; Calculate the basic travel cost across each edge based on the edge length and the robot's moving speed; Calculate the historical passage cost of each edge based on the congestion status of each edge within a preset historical time period. Based on the traffic information, the current congestion level of each side is identified, and the real-time traffic cost of each side is calculated. The real-time passage cost is updated based on the update time of the robot side.
3. The method of claim 2, wherein, The step of using the traffic information to establish a probabilistic model for predicting traffic flow for each robot, in order to obtain traffic flow prediction results, includes: The probability model is established based on the probability distribution of each robot moving normally at different times and at different nodes, and the probability distribution of each robot colliding at different times and at different nodes. The current operating status of each robot at each node is determined based on the position, path, and direction of each robot in the passage information; the operating status includes the probability of a robot colliding at each node and the probability of a robot moving normally at each node. Based on the operational status and the probability model, the probability distribution of each robot at different locations in the target area within a preset time period is predicted, and the traffic flow prediction result of the target area is obtained.
4. A multi-robot path planning method, wherein, include: Receive robot attribute information, corresponding toll costs, and traffic flow prediction results from the central manager, and obtain the target task; Based on the target task, the toll cost, and the traffic flow prediction results, a route plan is established for the target task; Obtain information about other robots within the target area and corresponding traffic information; Using the path planning and traffic information from the other robots, future travel costs are estimated; Based on the future travel cost and the travel cost fed back by the central manager, calculate the comprehensive cost of the robot's travel path from the starting point to the target point; Based on the comprehensive cost, the optimal travel route is determined.
5. The method of claim 4, wherein, The method further includes: Robots that meet preset conditions in the passageway are added as dynamic obstacles to a dynamic obstacle list; wherein, the preset conditions are determined according to the task priority of each robot; and the task priority is determined according to the time when each robot acquires the task. The travel path from the origin to the destination is replanned based on the future travel costs and the dynamic obstacle list.
6. The method of claim 4 or 5, wherein, The method further includes: When traveling to the target point according to the travel path, avoid the target robot passing through the target point.
7. A multi-robot path planning apparatus, wherein, include: The first acquisition module is used to acquire the passage information of each robot in the target area; The first determining module is used to determine the passage cost of each robot based on the passage information; The first module is used to establish a probabilistic model for predicting traffic flow for each robot using the traffic information, so as to obtain the traffic flow prediction result. The feedback module is used to respond to the robot's request information and provide the robot with the robot's attribute information and corresponding travel costs and traffic flow prediction results, so that the robot can calculate the future travel costs of passing through different locations in the target area based on the traffic flow prediction results, and plan the travel path from the starting point to the target point based on the future travel costs. The access information includes at least one of the following: the robot's identifier, location, direction, and path.
8. A multi-robot path planning apparatus, wherein, include: The receiving module is used to receive the robot's attribute information, corresponding passage costs, and traffic flow prediction results from the central manager, and to obtain the target task. The second module is used to establish a path plan for the target task based on the target task, the toll cost, and the traffic flow prediction results. The second acquisition module is used to acquire information about other robots within the target area and the corresponding traffic information; The estimation module is used to estimate future travel costs using the path planning and traffic information from the other robots; The calculation module is used to calculate the comprehensive cost of the robot's travel path from the starting point to the target point based on the future travel cost and the travel cost fed back by the central manager. The second determining module is used to determine the optimal travel path based on the comprehensive cost.
9. A multi-robot path planning system, wherein, include: Central manager and multiple robots; The central manager is configured to perform the method as described in any one of claims 1-3; The robot is used to perform the method as described in any one of claims 4-6.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-6.