Method for resolving dynamic conflicts of multiple agvs based on process priority and related components

Abstract

The application provides a multi-AGV dynamic conflict resolution method based on process priority and related components, and the technical scheme points are as follows: collecting and standardizing the process priority and load level of AGV; predicting the conflict risk based on a topological road network model; when the risk exceeds a threshold value, the process priority and the load level are fused to calculate the comprehensive traffic right to determine the traffic attribution; accordingly, the low-priority AGV is executed to wait for avoidance instructions or local re-planning to resolve the conflict; finally, the path is updated and the cycle is executed. Compared with the prior art, the application reduces the risk that the AGV on the key process path is blocked at the key node or road section, and improves the smoothness of the key process path of the production line.

Description

Technical Field

[0001] This invention relates to the field of automated guided vehicle (AGV) technology, specifically to a method for resolving dynamic conflicts among multiple AGVs based on process priority and related components. Background Technology

[0002] In intelligent manufacturing and flexible production systems, Automated Guided Vehicles (AGVs) are widely used in material handling and process coordination. When multiple AGVs operate collaboratively in the same physical space, operational conflicts are highly likely to occur due to limited road network resources (such as intersections and road segments), meaning multiple AGVs attempt to occupy the same node or road segment simultaneously. Existing technologies typically use rules based on task urgency, first-come-first-served, or fixed priorities to resolve conflicts. This can lead to AGVs performing critical process paths being blocked by AGVs performing ordinary material handling tasks, thus disrupting production line rhythm and affecting overall production efficiency. Summary of the Invention

[0003] The purpose of this invention is to provide a method and related components for resolving dynamic conflicts among multiple AGVs based on process priority, thereby reducing the risk of AGVs being blocked at critical nodes or sections on critical process paths and improving the smoothness of critical process paths on the production line.

[0004] In a first aspect, the invention provides a method for dynamic conflict resolution of multiple AGVs based on process priority, comprising: S1, collecting AGV status data and standardizing the status data to generate standardized status data, wherein the standardized status data includes standardized process priority and standardized load level; S2, based on a topology network model and the standardized status data, predicting the probability of AGV conflict at nodes or road segments, and generating a conflict resolution trigger signal when the probability of conflict exceeds a preset threshold; S3, in response to the conflict resolution trigger signal, based on the standardized process priority and the standardized load level, the method generates a conflict resolution trigger signal. S1) Describe the standardized load level, calculate the comprehensive right-of-way priority of the corresponding AGV, and determine the right-of-way ownership; S4) Based on the right-of-way ownership, issue a wait-and-avoidance instruction to the low-priority AGV, and determine whether there is a deadlock risk. If so, perform local replanning on the low-priority AGV to generate the optimal resolution solution; S5) Based on the optimal resolution solution, update the AGV's travel path, generate updated path information, and send the updated path information to the scheduling system; S6) Repeat steps S1 to S5 to dynamically monitor and resolve conflicts during AGV operation.

[0005] According to one embodiment of the present invention, the step of collecting AGV status data and standardizing the status data to generate standardized status data includes: collecting real-time position coordinates, driving speed, and task execution status reported by the AGV through onboard sensors, and obtaining corresponding process priority parameters and rated load parameters from a scheduling system or manufacturing execution system; mapping the process priority parameters to a first preset value range according to preset standardization rules to obtain the standardized process priority; calculating the ratio of real-time load to the rated load, and mapping the ratio to a second preset value range to obtain the standardized load level; and integrating the real-time position coordinates, driving speed, task execution status, standardized process priority, and standardized load level to generate the standardized status data.

[0006] According to one embodiment of the present invention, the step of performing conflict prediction based on a topology network model and standardized state data, calculating the probability of AGV conflict at nodes or road segments, and generating a conflict resolution trigger signal when the probability of conflict exceeds a preset threshold, includes: pre-constructing the topology network model based on the physical layout of the production site; wherein the topology network model defines nodes, road segments, and traffic rules; extracting the distance and travel speed of the AGV to potential conflict nodes from the standardized state data, and obtaining the remaining time occupied by the potential conflict nodes from the scheduling system; substituting the distance, travel speed, and remaining time into a conflict risk judgment formula based on the Sigmoid function to calculate the probability of conflict; determining whether the probability of conflict is greater than or equal to a preset probability threshold, and if so, generating the conflict resolution trigger signal containing a conflict node identifier and an AGV identifier.

[0007] According to one embodiment of the present invention, the step of calculating the comprehensive right-of-way priority of an AGV based on the standardized process priority and the standardized load level in response to the conflict resolution trigger signal, and determining the right-of-way allocation, includes: extracting the standardized process priority and the standardized load level of the corresponding AGV in response to the conflict resolution trigger signal; obtaining a pre-configured process priority weight coefficient and a load level weight coefficient, wherein the process priority weight coefficient is greater than the load level weight coefficient; substituting the standardized process priority, the standardized load level, the process priority weight coefficient, and the load level weight coefficient into a comprehensive right-of-way calculation model to calculate the comprehensive right-of-way priority of the corresponding AGV, wherein the comprehensive right-of-way calculation model is used to perform a weighted summation calculation; comparing the comprehensive right-of-way priorities of all AGVs, and determining the AGV with the highest priority value as having priority right-of-way.

[0008] According to one embodiment of the present invention, the step of issuing a wait-and-avoidance instruction to low-priority AGVs based on the right-of-way allocation and determining whether there is a deadlock risk, and if so, performing local replanning on the low-priority AGVs to generate an optimal resolution scheme, includes: determining high-priority AGVs and low-priority AGVs based on the right-of-way allocation; issuing a wait-and-avoidance instruction to the low-priority AGVs and constructing a directed graph of all current AGVs using a resource allocation graph method to determine whether executing the wait-and-avoidance instruction will cause a system deadlock; if no deadlock risk is determined, the wait-and-avoidance instruction is taken as the optimal resolution scheme; if a deadlock risk is determined, a search algorithm is used to search for feasible detour paths around the conflict area, and the comprehensive cost of each path is calculated based on the new path length, task delay time, and new path conflict risk probability; the path with the minimum comprehensive cost is selected as the optimal resolution scheme.

[0009] According to one embodiment of the present invention, the step of updating the AGV's travel path according to the optimal resolution scheme, generating updated path information, and sending the updated path information to the scheduling system includes: parsing the optimal resolution scheme; if the optimal resolution scheme is a waiting avoidance instruction, generating path information including pause points; if the optimal resolution scheme is a detour path, replacing the AGV's original travel path with the coordinate sequence of the detour path to generate updated path information; recalculating and updating the expected occupancy time windows of relevant nodes and road segments in the topology network model according to the updated path information to form global path occupancy information; and sending the updated path information and global path occupancy information to the scheduling system and the AGV through a communication interface.

[0010] According to one embodiment of the present invention, the cyclic execution steps S1 to S5, for dynamically monitoring and resolving conflicts during AGV operation, include: after completing one conflict resolution, receiving new AGV driving status data fed back from the scheduling system and using it as the input data source for a new cycle; continuously collecting AGV status data at a preset sampling frequency and dynamically updating the standardized status data; and based on the updated standardized status data, cyclically executing conflict prediction, right-of-way determination, and conflict resolution steps to achieve adaptive closed-loop scheduling.

[0011] Secondly, the present invention also provides a multi-AGV dynamic conflict resolution device based on process priority, comprising: a status acquisition module, used to acquire AGV status data and standardize the status data to generate standardized status data, the standardized status data including standardized process priority and standardized load level; a conflict prediction module, used to predict the probability of AGV conflict at nodes or road segments based on a topology network model and the standardized status data, and generate a conflict resolution trigger signal when the conflict probability exceeds a preset threshold; and a right-of-way determination module, used to respond to the conflict resolution trigger signal and, based on the standardized process priority and the standardized load level, determine the right-of-way based on the standardized process priority and the standardized load level. The system comprises a standardized load level, a comprehensive right-of-way priority for the corresponding AGV, and a right-of-way allocation module. A conflict resolution module issues wait-and-avoidance instructions to low-priority AGVs based on the right-of-way allocation and determines if deadlock risk exists. If so, it performs local replanning on the low-priority AGVs to generate an optimal resolution solution. A path update module updates the AGV's travel path based on the optimal resolution solution, generates updated path information, and sends the updated path information to the scheduling system. A closed-loop scheduling module cyclically controls the above modules to execute the corresponding steps, dynamically monitoring and resolving conflicts during AGV operation.

[0012] Thirdly, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the process priority-based multi-AGV dynamic conflict resolution method of the above embodiments.

[0013] Fourthly, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the multi-AGV dynamic conflict resolution method based on process priority described in the above embodiments.

[0014] The present invention produces at least the following beneficial effects: This invention achieves deep matching between scheduling logic and production logic and flexible and safe resolution process by integrating process priority and load level into a comprehensive right-of-way determination model, and adopting a combined resolution strategy of waiting and avoidance, deadlock determination, and local replanning. This reduces the risk of AGVs being blocked at key nodes or sections on critical process paths and improves the smoothness of critical process paths on the production line. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a multi-AGV dynamic conflict resolution method based on process priority, according to an embodiment of the present invention. Figure 2 yes Figure 1 A flowchart illustrating a specific implementation method of step S1; Figure 3 yes Figure 1 A flowchart illustrating a specific implementation method of step S2; Figure 4 yes Figure 1 A flowchart illustrating a specific implementation method of step S3; Figure 5 yes Figure 1 A flowchart illustrating a specific implementation of step S4; Figure 6 yes Figure 1 A flowchart illustrating a specific implementation of step S5; Figure 7 yes Figure 1 A flowchart illustrating a specific implementation of step S6; Figure 8 This is a schematic block diagram of a multi-AGV dynamic conflict resolution device based on process priority, according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] It should be understood that, when used in this specification and the appended claims, the terms “comprising” and “including” indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more of its features, integrals, steps, operations, elements, components and / or collections thereof.

[0019] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0020] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0021] To facilitate the description of this application, the following concepts related to this application are introduced.

[0022] Automated Guided Vehicle (AGV): A transport vehicle equipped with electromagnetic or optical automatic guidance devices, capable of traveling along a prescribed guidance path, and possessing safety protection and various transfer functions. In this application, the AGV is the main entity performing material handling tasks, and its status data is the basis for dynamic conflict prediction and resolution.

[0023] Process priority: refers to a numerical value obtained by quantifying and classifying the importance, urgency, or core nature of production processes according to production process rules. In this application, process priority is defined by the Manufacturing Execution System (MAS) and is used to prioritize high-value, high-cycle production processes during conflict resolution.

[0024] Load rating: This refers to a standardized classification based on the ratio of the AGV's real-time cargo weight to its rated load capacity. In this application, the load rating serves as a supplementary basis for determining right-of-way, aiming to optimize the utilization of road network resources and prevent empty vehicles from blocking fully loaded vehicles.

[0025] Topology network model: An environmental abstraction model based on graph theory, which abstracts the physical path network of the production site into a set of nodes and edges. In the scheme of this application, nodes represent key locations such as intersections and workstations, and edges represent paths connecting nodes. The topology network model is used to predict spatiotemporal conflicts of AGVs at specific nodes or road segments.

[0026] Deadlock risk: In a multi-tasking or multi-process system, two or more execution units become indefinitely blocked because they are waiting for each other to release resources. In this application, it refers to a deadlock of circular waiting and resource mutual exclusion that may occur when multiple AGVs execute a wait-and-avoidance strategy.

[0027] Local replanning refers to recalculating and optimizing only the path in a local area when dynamic obstacles or conflicts are detected, based on global path planning. In this application, it is used to find the least costly detour path for low-priority AGVs under deadlock risk.

[0028] This application provides a method for dynamic conflict resolution of multiple AGVs based on process priority, referring to... Figure 1 ,include: S1. Collect the status data of the AGV and standardize the status data to generate standardized status data, which includes standardized process priority and standardized load level.

[0029] The purpose of this step is to acquire the real-time and comprehensive operating status and task attributes of all AGVs, and to eliminate the dimensional differences of different indicators through standardization, providing a foundation for subsequent unified calculations and decisions. Specifically, the real-time position coordinates and travel speed are collected through the positioning module (such as UWB, laser SLAM) and speed sensor on the AGV; the task execution status, such as in progress or nearing completion, is obtained through the task queue of the scheduling system. At the same time, the raw value of the process priority corresponding to the current task is obtained from the manufacturing execution system interface, and the rated load is obtained from the AGV management database. According to preset rules, the process priority is mapped to the range of [1, 10], and the ratio of real-time load to rated load is mapped to the range of [1, 5], generating standardized process priorities and load levels, which are then integrated with position, speed, and other data.

[0030] S2. Based on the topology network model and the standardized state data, predict the probability of AGV conflict at nodes or road segments. When the probability of conflict exceeds a preset threshold, generate a conflict resolution trigger signal.

[0031] The purpose of this step is to anticipate potential conflicts in advance, transforming passive response into proactive intervention. Specifically, the system loads a pre-constructed topology network model based on the workshop layout. This model defines all nodes (such as work points and intersections) and road segments, along with their capacity. Combining the standardized status data generated in S1, the estimated time for each AGV to reach a potential conflict point (such as an intersection) is calculated. Using a conflict risk assessment formula, the difference between the estimated arrival time and the remaining time that the point is occupied by other AGVs is converted into a conflict probability between 0 and 1. When this probability exceeds a preset threshold (e.g., 0.8), a trigger signal containing the conflict location and the corresponding AGV number is immediately generated, initiating the resolution process.

[0032] S3. In response to the conflict resolution trigger signal, calculate the comprehensive right-of-way priority of the corresponding AGV based on the standardized process priority and the standardized load level, and determine the right-of-way ownership.

[0033] The purpose of this step is to fairly and efficiently determine which AGV should have priority passage when a conflict occurs, based on a set of rules that integrate production logic and efficiency optimization. Specifically, upon receiving a conflict resolution trigger signal, the system extracts the standardized process priority and standardized load level of all corresponding AGVs. It obtains preset weighting coefficients, where the process priority weighting coefficient (e.g., 0.8) is greater than the load level weighting coefficient (e.g., 0.2). These two parameters are then substituted into the comprehensive passage right calculation model (i.e., a weighted summation formula) to calculate a comprehensive passage right priority value for each relevant AGV. By comparing these values, the AGV with the highest value is determined to have priority passage.

[0034] S4. Based on the right-of-way allocation, issue a wait-and-avoidance instruction to the low-priority AGV and determine whether there is a deadlock risk. If so, perform local replanning on the low-priority AGV to generate the optimal solution.

[0035] The purpose of this step is to implement a flexible and safe conflict resolution strategy based on the right-of-way determination. Specifically, after identifying high-priority and low-priority AGVs, an instruction is first issued to the low-priority AGV, instructing it to pause and wait in the safe buffer zone before the conflict point. Simultaneously, the system uses graph theory to construct a directed graph of resource waiting for all AGVs, analyzing whether a circular waiting chain exists. If no deadlock risk is determined, waiting and avoiding is the optimal solution. If a deadlock risk is determined, a local replanning is performed on the low-priority AGV. Between its current position and the target point, all feasible detour paths are searched around the conflict area, and the comprehensive cost of each path is calculated, taking into account path length, additional delay time, etc., selecting the path with the lowest cost as the optimal resolution solution.

[0036] S5. Based on the optimal solution, update the AGV's travel path, generate updated path information, and send the updated path information to the scheduling system.

[0037] The purpose of this step is to translate the decision-making scheme into executable instructions and synchronously update the global path view. Specifically, if the optimal solution is to wait and avoid obstacles, path information including pause coordinates and resumption conditions is generated. If it is a detour path, the original path segments of the AGV are replaced with the new path coordinate sequence. Based on the updated path, the time windows for the occupation of relevant nodes and road segments in the topology network model are recalculated and updated to form a globally consistent path occupancy time sequence table. Finally, the updated individual path instructions and global occupancy information are sent to the scheduling system and all relevant AGVs via industrial Ethernet or wireless network.

[0038] S6. Repeat steps S1 to S5 to dynamically monitor and resolve conflicts during AGV operation.

[0039] The purpose of this step is to achieve continuity and adaptability in the scheduling process, forming a closed-loop control. Specifically, after a conflict is resolved, the system does not stop but immediately receives the latest AGV status data from the scheduling system as input for the next cycle. The system continuously collects and standardizes the AGV status at a fixed high frequency (e.g., 10Hz) and, based on the latest topology network model occupancy, cyclically executes conflict prediction, right-of-way determination, and conflict resolution decisions, thereby achieving 24 / 7 uninterrupted dynamic monitoring and rapid conflict resolution of the multi-AGV system's operating status.

[0040] The embodiments of this application employ the above method, which produces at least the following beneficial effects: By integrating process priority and load level into a comprehensive right-of-way determination model, and employing a combined resolution strategy of waiting and avoidance, deadlock detection, and local replanning, this method achieves deep matching between scheduling logic and production logic, as well as flexible and safe resolution processes. This approach addresses the problem of high-priority tasks being blocked or the system deadlocking due to a single strategy. It improves the smoothness of critical production processes and enhances the efficiency of multi-AGV collaborative operation.

[0041] By collecting and standardizing multi-dimensional status data such as process priority and load level of AGVs in real time, a comprehensive and quantitative perception of AGV operating status is achieved. This method solves the problem that traditional scheduling systems only focus on position and speed, while ignoring task attributes and load information. This method provides a precise data foundation for intelligent decision-making based on production processes.

[0042] By predicting conflict probability based on a topological road network model and the sigmoid function, this method achieves early, quantitative warnings of potential conflicts. It addresses the response delays caused by existing technologies that often only address conflicts after they occur, thus shortening the conflict resolution response time and shifting from post-event processing to pre-event prevention.

[0043] This method can be applied to multi-AGV collaborative material delivery scenarios in automobile manufacturing final assembly workshops. Given the strict cycle times of the final assembly line and the unavoidable delays in the delivery of high-priority components such as seats and engines, this method ensures priority passage for these critical AGVs through process prioritization and prevents them from being blocked at intersections by ordinary material delivery AGVs through dynamic conflict prediction and resolution, thereby improving the efficiency of the final assembly line.

[0044] This method can be applied to flexible connection scenarios between SMT production lines and assembly lines in the 3C electronics industry. In this scenario, AGVs are diverse in model, have fast task cycles, and numerous intersecting paths exist. This method can optimize the obstacle avoidance strategy of idle AGVs based on real-time load and quickly resolve sudden conflicts caused by temporary equipment placement or personnel movement through local replanning, improving the timeliness and safety of precision material handling.

[0045] In some embodiments, reference Figure 2 Step S1 includes: S11: Collect the real-time position coordinates, driving speed and task execution status reported by the AGV through the on-board sensors, and obtain the corresponding process priority parameters and rated load parameters from the scheduling system or manufacturing execution system. S12: According to the preset standardization rules, the process priority parameter is mapped to a first preset numerical range to obtain the standardized process priority; S13: Calculate the ratio of the real-time load to the rated load, and map the ratio to a second preset value range to obtain the standardized load level; S14: Integrate the real-time location coordinates, driving speed, task execution status, standardized process priority, and standardized load level to generate the standardized status data; The specific implementation steps are as follows: The first step is multi-source data acquisition: Real-time position (x, y) and speed v provided by the AGV's onboard encoder are read via the CAN bus; the current task ID and status are queried through the scheduling system's task management interface, such as heading to workstation A; the corresponding process priority raw value P_raw is queried based on the task ID via the manufacturing execution system's API, where the process priority raw value P_raw can be any of high, medium, low, or numerical values; the rated load C_max for this AGV model is read from the AGV asset database. The second step is process priority standardization: Preset mapping rules, for example: if P_raw is high, it is mapped to 10; if P_raw is medium, it is mapped to 6; if P_raw is low, it is mapped to 3; if P_raw itself is a value in the range [1, 100], it is linearly scaled to the range [1, 10]. This yields the standardized process priority P_proc. The third step is load level standardization: The real-time load C_current is obtained using the vehicle's weight sensor. The load ratio r = C_current / C_max is calculated. Preset mapping rules: If r ≥ 0.8, the standardized load level P_load is assigned a value of 5 (full load); if r < 0.2, it is assigned a value of 1 (no load); if 0.2 ≤ r < 0.8, it is assigned a value within the (1,5) interval through linear interpolation. The fourth step is data integration and packaging: fields such as location, speed, task status, P_proc, and P_load are encapsulated according to a predefined data structure (such as a JSON object or a specific class instance), and timestamped to generate a complete standardized status data packet, which is then sent to the subsequent processing module.

[0046] Through the standardized data acquisition and processing procedures described above, step S1 achieves the unification and normalization of heterogeneous raw data. Step S1 solves the problem that data from different sources with different formats and units cannot be directly used for calculation, thus providing reliable and consistent input for subsequent conflict prediction and right-of-way calculation based on a unified mathematical model, thereby improving the accuracy of the input data.

[0047] In some embodiments, reference Figure 3 Step S2 includes: S21: The topology network model is pre-constructed based on the physical layout of the production site; wherein, the topology network model defines nodes, road segments and traffic rules; S22: Extract the distance and speed of the AGV to the potential conflict node from the standardized state data, and obtain the remaining time that the potential conflict node is occupied from the scheduling system; S23: Substitute the distance, the driving speed, and the remaining time into the conflict risk determination formula based on the Sigmoid function to calculate the probability of the conflict occurring; S24: Determine whether the probability of the conflict occurring is greater than or equal to a preset probability threshold. If so, generate the conflict resolution trigger signal containing the conflict node identifier and the AGV identifier.

[0048] The specific implementation steps are as follows: The first step is model loading: During system initialization, a topology network model file describing the workshop layout is loaded. This file abstracts the workshop as a set of nodes N (each node has a unique ID and coordinates) and a set of edges E (connecting two nodes and having a length attribute). This topology network model also defines the traffic capacity of each node / segment (usually 1). The second step is to obtain key parameters: For an AGV_i that is heading towards a node_k, extract the straight-line distance d_i from its current position to the node_k and its current speed v_i from its standardized state data. Simultaneously, query the global path schedule maintained by the scheduling system to find the AGV_j that is currently occupying or about to occupy the node_k, and calculate the remaining time t_j for AGV_j to leave the node_k. The third step is probability calculation: For the current AGV (denoted as AGV_i) and other AGVs that may conflict with it (denoted as AGV_j), substitute the distance d_i from AGV_i to the potential conflict node, the current travel speed v_i of AGV_i, and the remaining time t_j for AGV_j to occupy the conflict node into the conflict risk determination formula based on the Sigmoid function: ; Where d_i is in meters (m), v_i is in meters per second (m / s), and t_j is in seconds (s). d_i / v_i represents the estimated remaining time for AGV_i to reach the conflict node, and t_j represents the remaining time for the conflict node to be released by AGV_j. The output value of this formula is between 0 and 1, with the following specific meanings: When d_i / v_i = t_j (AGV arrival time equals conflict node release time), the exponent is 0, R_conflict = 0.5, indicating a moderate probability of conflict; When d_i / v_i > t_j (AGV arrives first and needs to wait), R_conflict > 0.5, and the larger the difference (i.e. the earlier AGV_i arrives), the closer R_conflict is to 1, indicating that the probability of conflict is extremely high. When d_i / v_i < t_j (the node has been released when the AGV arrives), R_conflict < 0.5, and the larger the difference, the closer R_conflict is to 0, indicating that the probability of conflict is extremely low.

[0049] This formula enables continuous and smooth quantification of conflict probabilities.

[0050] Step 4, Threshold Determination and Triggering: Set the probability threshold to R_threshold = 0.8. The calculated... If the value is greater than or equal to 0.8, it is determined that a high-risk conflict is about to occur. The system immediately generates a trigger signal, which contains at least the conflict node ID (Node_k) and the corresponding AGV IDs (AGV_i and AGV_j), and sends it to the priority determination module.

[0051] By employing a conflict risk assessment formula built using the Sigmoid function, step S2 achieves continuous, non-linear quantification of conflict probability, replacing the traditional binary judgment. Step S2 addresses the insensitivity of the fixed-time threshold method to boundary conditions, thereby making conflict prediction more accurate and closely aligned with the dynamic characteristics of AGV movement, enabling earlier identification of potential conflicts.

[0052] In some embodiments, reference Figure 4 Step S3 includes: S31: In response to the conflict resolution trigger signal, extract the standardized process priority and the standardized load level of the corresponding AGV; S32: Obtain the pre-configured process priority weight coefficient and load level weight coefficient of the system, wherein the process priority weight coefficient is greater than the load level weight coefficient; S33: Substitute the standardized process priority, standardized load level, process priority weight coefficient, and load level weight coefficient into the comprehensive access right calculation model to calculate the comprehensive access right priority of the corresponding AGV, wherein the comprehensive access right calculation model is used to perform weighted summation calculation; S34: Compare the overall right-of-way priority of all AGVs and determine the AGV with the highest priority value to obtain priority right-of-way.

[0053] The specific implementation steps are as follows: The first step is data extraction: Based on the list of AGV IDs (e.g., AGV_A, AGV_B) in the conflict resolution trigger signal, extract the process priority P_proc_A and load level P_load_A of AGV_A, as well as P_proc_B and P_load_B of AGV_B from the latest standardized status data pool. The second step is to obtain the weights: read the preset weight coefficients from the system configuration file; The third step is weighted calculation: For each AGV, the comprehensive right-of-way calculation model is applied: P_total = α · P_proc + β · P_load. In this formula, α is the process priority weight, ranging from [0.7, 0.9], ensuring that the production process priority is the core criterion for judgment; β is the load level weight, ranging from [0.1, 0.3], serving as a supplementary optimization basis, and satisfying α + β = 1 to achieve weight normalization.

[0054] For example, AGV_A: P_total_A = 0.8×8 + 0.2×5 = 6.4 + 1.0 = 7.4; AGV_B: P_total_B = 0.8×6 + 0.2×1 = 4.8 + 0.2 = 5.0.

[0055] Step 4, attribution determination: Compare the calculated P_total values. The higher the P_total value, the higher the overall right-of-way priority. For example, in this case, P_total_A (7.4) > P_total_B (5.0), therefore AGV_A is determined to have priority right-of-way, while AGV_B is a low-priority AGV that needs to give way or detour.

[0056] By constructing a comprehensive right-of-way calculation model with process priority as the core weight, step S3 deeply embeds production scheduling logic into real-time traffic decision-making. Step S3 solves the problem of the disconnect between traditional conflict resolution rules based on first-come-first-served or shortest path and actual production needs, thereby ensuring that AGVs performing critical processes can obtain priority right-of-way and improving the overall production line efficiency.

[0057] In some embodiments, reference Figure 5 Step S4 includes: S41: Determine the high-priority AGV and low-priority AGV based on the right-of-way allocation; S42: Issue a wait-and-avoid command to the low-priority AGV, and construct a directed graph of all current AGVs using the resource allocation graph method to determine whether executing the wait-and-avoid command will cause a system deadlock. S43: If it is determined that there is no risk of deadlock, then the waiting and avoidance instruction shall be taken as the optimal solution. S44: If a deadlock risk is determined, a search algorithm is used to search for feasible detour paths around the conflict area, and the comprehensive cost of each path is calculated based on the length of the new path, the task delay time, and the probability of conflict risk of the new path. S45: Select the path with the lowest overall cost as the optimal solution.

[0058] The purpose of this step is to elaborate on the decision-making process for a combined conflict resolution strategy. This step aims to provide a safer alternative when simply waiting and avoiding conflict may lead to systemic problems. The specific implementation steps are as follows: The first step, initial instruction issuance and deadlock detection: If AGV_B is determined to be a low-priority AGV, the system issues an instruction to it to stop and wait at a preset buffer point a certain distance from the conflict node. Simultaneously, the system constructs a resource allocation graph, treating all AGVs as processes and their currently occupied and requested path nodes as resources. By detecting whether loops exist in the graph, it determines whether AGV_B's waiting will lead to a circular waiting chain of multiple AGVs. The second step is handling the case without deadlock: If the deadlock detection algorithm (such as loop detection) returns a result indicating no loop, it means that the current waiting and avoidance is safe. The system will determine the sequence of instructions for AGV_B waiting at the buffer point and for AGV_A to pass through as the optimal solution for resolving this conflict. The third step is replanning in case of deadlock: If a loop is detected, a deadlock risk is identified. The system immediately cancels the wait command for AGV_B and initiates local replanning instead. Starting from the current position of AGV_B and taking a safe point after the conflict node in its original path as the temporary endpoint, the system searches for all feasible paths that avoid the conflict node in the topology network model. Step 4, Path Cost Evaluation and Selection: For each feasible detour path found, calculate its comprehensive cost: Cost_replan = λ·L_new + μ·T_delay + γ·R_new. Where: L_new is the length of the new detour path, in meters (m). T_delay is the task delay time caused by detour, in seconds (s); R_new represents the probability of conflict risk for the new path, determined by the conflict risk assessment formula: The calculation yielded the result. λ, μ, and γ are weighting coefficients that satisfy λ + μ + γ = 1, used to balance the importance of path length, time delay, and conflict risk. For example, λ = 0.6, μ = 0.3, and γ = 0.1 can be set, which focuses more on optimizing path length; or λ = 0.4, μ = 0.4, and γ = 0.2 can be set, which more evenly considers the three indicators. The path with the smallest Cost_replan value is selected as the optimal detour path, and the optimal solution is generated based on this.

[0059] By introducing deadlock risk assessment and switching to local replanning when necessary, step S4 constructs a two-layer decision-making mechanism that combines safety and flexibility. Step S4 solves the problem that a single waiting strategy can easily lead to system-level deadlock paralysis in complex road networks, thus enabling the scheduling system of multiple AGVs to maintain high overall activity and reliability while efficiently resolving conflicts.

[0060] In some embodiments, reference Figure 6 Step S5 includes: S51: Analyze the optimal solution. If the optimal solution is to wait for an avoidance command, then generate path information containing the pause point. S52: If the optimal solution is a detour path, then replace the original travel path of the AGV with the coordinate sequence of the detour path to generate updated path information. S53: Based on the updated path information, recalculate and update the estimated occupancy time windows of relevant nodes and road segments in the topology network model to form global path occupancy information; S54: The updated path information and global path occupancy information are sent to the scheduling system and AGV through the communication interface.

[0061] The specific implementation steps are as follows: The first step, solution analysis and path generation: If the solution is to wait for a collision, a path containing two key points is generated for AGV_B: 1) Immediately decelerate and travel to the specified buffer point coordinates; 2) Stop at the buffer point until a recovery command is received. If the solution is to detour, a new path coordinate sequence, such as [x1,y1], [x2,y2],..., is planned for AGV_B; completely replacing the segment of its original path from the current position to the re-merging point after bypassing the conflict zone. The second step is a global time window update: Based on the new path of AGV_B, whether it pauses or detours, the scheduling system recalculates the arrival and departure times of each node and segment on the path. These new time windows are then used to update the occupancy schedules of the corresponding elements in the topology network model. For example, the time slot originally allocated to the conflicting node Node_k by AGV_B is now released, while new nodes on the detour path are added to the occupancy plan. The third step is instruction issuance and synchronization: via a wireless network, such as Wi-Fi 6 or a 5G private network, the updated detailed path instructions for AGV_B, including coordinates and speed suggestions, are issued to its onboard controller. Simultaneously, the updated global path occupancy information table is broadcast to all relevant modules in the scheduling system, such as the path planner and monitoring interface, ensuring that the entire system has a consistent understanding of the latest road network resource occupancy status.

[0062] By synchronously updating the individual AGV paths and the global road network occupancy time windows, step S5 achieves consistency maintenance between local decisions and the global state. Step S5 resolves new planning conflicts caused by information asynchrony, such as robbing Peter to pay Paul, thus enabling the orderly resolution of high-frequency dynamic conflicts under a stable and consistent global view, improving the overall coordination efficiency of the scheduling system.

[0063] In some embodiments, reference Figure 7 Step S6 includes: S61: After completing a conflict resolution, receive the new AGV driving status data fed back from the scheduling system and use it as the input data source for the new cycle; S62: Continuously collect AGV status data at a preset sampling frequency and dynamically update the standardized status data; S63: Based on the updated standardized state data, iteratively executes conflict prediction, right-of-way determination, and conflict resolution steps to achieve adaptive closed-loop scheduling.

[0064] The specific implementation steps are as follows: The first step is data feedback: Once AGV_B starts executing a new path (waiting or detouring), its onboard sensors continuously report new position and speed data. This data is aggregated by the scheduling system and fed back as the latest status to step S1 of this method. The second step is continuous data acquisition and updating: The system executes step S1 at a fixed high frequency (e.g., 10 times per second), continuously acquiring the latest status of all AGVs and generating new standardized status data packets to overwrite the old data. This ensures that the real-time position, speed, and process priority of AGV_B, which may change due to task updates, are captured in real time. The third step is closed-loop execution: using the updated standardized state data as input, the system executes steps S2, S3, S4, and S5 again. This process repeats continuously, enabling the system to dynamically respond to any events that may cause conflict, such as AGV position changes, new task insertions, and temporary obstacles, forming a complete closed loop of perception, decision-making, execution, and feedback.

[0065] By establishing a closed-loop process that runs continuously at a fixed frequency, step S6 achieves perpetual monitoring and dynamic optimization of the multi-AGV system's operating status. Step S6 solves the monitoring blind spots and response lag problems that exist in intermittent or trigger-based scheduling methods in continuous operation scenarios, thus providing continuous, stable, and adaptive conflict resolution guarantees for high-paced, high-concurrency production logistics scenarios.

[0066] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0067] Reference Figure 8 This invention provides a multi-AGV dynamic conflict resolution device 100 based on process priority, comprising: The status acquisition module 101 is used to: S1. acquire the status data of the AGV, and standardize the status data to generate standardized status data, which includes standardized process priority and standardized load level. The conflict prediction module 102 is used to predict the probability of conflict occurring at nodes or road segments of AGV based on the topology road network model and the standardized state data in S2. When the probability of conflict occurs exceeds a preset threshold, a conflict resolution trigger signal is generated. The right-of-way determination module 103 is used in S3 to respond to the conflict resolution trigger signal, calculate the comprehensive right-of-way priority of the corresponding AGV based on the standardized process priority and the standardized load level, and determine the right-of-way ownership. The conflict resolution module 104 is used to: S4, issue a wait-and-avoidance instruction to the low-priority AGV according to the right-of-way ownership, and determine whether there is a deadlock risk. If there is, perform local replanning on the low-priority AGV to generate the optimal resolution solution. The path update module 105 is used in S5 to update the AGV's driving path according to the optimal solution, generate updated path information, and send the updated path information to the scheduling system. The closed-loop scheduling module 106 executes steps S1 to S5 in a loop to dynamically monitor and resolve conflicts during AGV operation.

[0068] Specific limitations regarding the process priority-based multi-AGV dynamic conflict resolution device 100 can be found in the limitations of the process priority-based multi-AGV dynamic conflict resolution method described above, and will not be repeated here. Each module in the aforementioned process priority-based multi-AGV dynamic conflict resolution device 100 can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0069] This invention also provides a computer device, which includes a processor, a memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface communicates with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of the process priority-based multi-AGV dynamic conflict resolution method provided in the above embodiments.

[0070] This invention also provides a computer-readable storage medium storing a computer program, which, when executed, can implement the functions or steps of the multi-AGV dynamic conflict resolution method based on process priority provided in the above embodiments.

[0071] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0072] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0074] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A dynamic conflict resolution method for multiple AGVs based on process priority, characterized in that, include: S1. Collect the status data of the AGV and standardize the status data to generate standardized status data, which includes standardized process priority and standardized load level. S2. Based on the topology network model and the standardized state data, predict the probability of AGV conflict at nodes or road segments. When the probability of conflict exceeds a preset threshold, generate a conflict resolution trigger signal. S3. In response to the conflict resolution trigger signal, calculate the comprehensive right-of-way priority of the corresponding AGV based on the standardized process priority and the standardized load level, and determine the right-of-way ownership. S4. Based on the right-of-way allocation, issue a wait-and-avoidance instruction to the low-priority AGV and determine whether there is a deadlock risk. If so, perform local replanning on the low-priority AGV to generate the optimal solution. S5. According to the optimal solution, update the AGV's travel path, generate updated path information, and send the updated path information to the scheduling system. S6. Repeat steps S1 to S5 to dynamically monitor and resolve conflicts during AGV operation.

2. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, The process of collecting AGV status data and standardizing the status data to generate standardized status data includes: Collect real-time location coordinates, driving speed, and task execution status reported by the AGV through on-board sensors, and obtain the corresponding process priority parameters and rated load parameters from the scheduling system or manufacturing execution system; According to the preset standardization rules, the process priority parameter is mapped to a first preset numerical range to obtain the standardized process priority; Calculate the ratio of the real-time load to the rated load, and map the ratio to a second preset value range to obtain the standardized load level; The standardized status data is generated by integrating the real-time location coordinates, driving speed, task execution status, standardized process priority, and standardized load level.

3. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, The method involves conflict prediction based on the topology network model and the standardized state data, calculating the probability of conflict occurrence for the AGV at nodes or road segments, and generating a conflict resolution trigger signal when the probability of conflict occurrence exceeds a preset threshold. This includes: The topology network model is pre-constructed based on the physical layout of the production site; wherein, the topology network model defines nodes, road segments and traffic rules; The distance and speed of the AGV to the potential conflict node are extracted from the standardized state data, and the remaining time occupied by the potential conflict node is obtained from the scheduling system. Substitute the distance, the driving speed, and the remaining time into the conflict risk determination formula based on the Sigmoid function to calculate the probability of the conflict occurring; Determine whether the probability of the conflict occurring is greater than or equal to a preset probability threshold. If so, generate the conflict resolution trigger signal containing the conflict node identifier and the AGV identifier.

4. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, In response to the conflict resolution trigger signal, the AGV's comprehensive right-of-way priority is calculated based on the standardized process priority and the standardized load level, and the right-of-way allocation is determined, including: In response to the conflict resolution trigger signal, the standardized process priority and the standardized load level of the corresponding AGV are extracted; Obtain the pre-configured process priority weight coefficient and load level weight coefficient of the system, wherein the process priority weight coefficient is greater than the load level weight coefficient; The standardized process priority, standardized load level, process priority weight coefficient, and load level weight coefficient are substituted into the comprehensive access right calculation model to calculate the comprehensive access right priority of the corresponding AGV. The comprehensive access right calculation model is used to perform weighted summation calculation. Compare the overall right-of-way priority of all AGVs, and determine the AGV with the highest priority value to obtain priority right-of-way.

5. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, The process involves issuing wait-and-avoidance instructions to low-priority AGVs based on the right-of-way allocation, determining whether there is a deadlock risk, and if so, performing local replanning on the low-priority AGVs to generate an optimal solution, including: Based on the right-of-way allocation, high-priority AGVs and low-priority AGVs are determined; A wait-and-avoid command is issued to the low-priority AGVs, and a directed graph of all current AGVs is constructed using a resource allocation graph method to determine whether executing the wait-and-avoid command will cause a system deadlock. If it is determined that there is no risk of deadlock, then the waiting and avoidance instruction is taken as the optimal solution. If a deadlock risk is identified, a search algorithm is used to search for feasible detour paths around the conflict area, and the comprehensive cost of each path is calculated based on the length of the new path, the task delay time, and the probability of conflict risk of the new path. The path with the lowest overall cost is selected as the optimal solution.

6. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, The step of updating the AGV's travel path according to the optimal resolution scheme, generating updated path information, and sending the updated path information to the scheduling system includes: The optimal solution is analyzed. If the optimal solution is to wait for an avoidance command, path information containing the pause point is generated. If the optimal solution is a detour path, then the coordinate sequence of the detour path replaces the original travel path of the AGV to generate updated path information. Based on the updated path information, the estimated occupancy time windows of relevant nodes and road segments in the topology network model are recalculated and updated to form global path occupancy information; The updated path information and global path occupancy information are sent to the scheduling system and AGV via the communication interface.

7. The multi-AGV dynamic conflict resolution method based on process priority according to claim 1, characterized in that, The cyclic execution steps S1 to S5 are used to dynamically monitor and resolve conflicts during AGV operation, including: After resolving a conflict, the system receives new AGV driving status data from the scheduling system and uses it as the input data source for the next cycle. The AGV's status data is continuously collected at a preset sampling frequency, and the standardized status data is dynamically updated. Based on the updated standardized state data, the system iteratively executes conflict prediction, right-of-way determination, and conflict resolution steps to achieve adaptive closed-loop scheduling.

8. A multi-AGV dynamic conflict resolution device based on process priority, characterized in that, include: The status acquisition module is used to collect the status data of the AGV and standardize the status data to generate standardized status data, which includes standardized process priority and standardized load level. The conflict prediction module is used to predict the probability of conflict occurring at nodes or road segments of AGV based on the topology road network model and the standardized state data. When the probability of conflict occurs exceeds a preset threshold, a conflict resolution trigger signal is generated. The right-of-way determination module is used to respond to the conflict resolution trigger signal, calculate the comprehensive right-of-way priority of the corresponding AGV based on the standardized process priority and the standardized load level, and determine the right-of-way ownership. The conflict resolution module is used to issue a wait-and-avoid command to low-priority AGVs based on the right-of-way ownership, and to determine whether there is a deadlock risk. If so, it performs local replanning on the low-priority AGVs to generate the optimal resolution solution. The path update module is used to update the AGV's travel path according to the optimal solution, generate updated path information, and send the updated path information to the scheduling system. The closed-loop scheduling module cyclically controls the above modules to execute the corresponding steps in order to dynamically monitor and resolve conflicts during AGV operation.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the multi-AGV dynamic conflict resolution method based on process priority as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the multi-AGV dynamic conflict resolution method based on process priority as described in any one of claims 1 to 7.

Images