An AGV path scheduling method based on spatial A-star algorithm and traffic control

Abstract

The present invention belongs to the field of AGV scheduling, and relates to an AGV path scheduling method based on a spatial A-star algorithm and traffic control. The method comprises: collecting current environmental data, identifying and marking key spatial elements on a map; using an improved spatial A-star algorithm to search for paths in the environment, and using the key spatial elements as key information in the search process to participate in path planning; setting multiple traffic control rules in path planning; when the AGV executes the travel path, monitoring the environment in real time, and replanning the path when encountering environmental changes; performing anomaly detection and collision prediction when the AGV executes the travel path, and adjusting the path of the AGV in real time according to the results of anomaly detection and collision prediction; the AGV continues to perform tasks according to the updated path until it reaches the destination. The AGV path scheduling method provided by the present invention adopts an improved spatial A-star algorithm, and significantly optimizes the path search process by introducing spatial information as an additional heuristic factor.

Description

Technical Field

[0001] The present invention belongs to the technical field of AGV scheduling, and in particular relates to an AGV path scheduling method based on a spatial A-star algorithm and traffic control. Background Art

[0002] With the rapid advancement of global industrial automation and intelligent manufacturing technology, automated guided vehicles (AGVs) have become an indispensable part of modern industrial production and logistics management. AGVs can autonomously complete tasks such as material handling and finished product transportation without direct human operation, significantly improving production efficiency and logistics efficiency while also reducing labor costs and potential human errors. As companies continue to increase their requirements for production efficiency and logistics response speed, the path planning technology of AGV systems is also facing new challenges and demands.

[0003] Learning-based planning methods, such as path planning algorithms based on machine learning or deep learning, can theoretically optimize path selection by learning from historical data on the environment and task execution, but in practice these methods usually require a large amount of data and training time to achieve the desired effect. In addition, learning-based methods may not always make the best decision when faced with new environments or situations that have never been encountered, especially in industrial environments where changes are complex and difficult to predict, which increases safety risks.

[0004] Traditional path planning methods, such as the classic A* search algorithm (i.e., A-star algorithm), have been widely used in many fields due to its simplicity and effectiveness. The A* algorithm searches for a path with the lowest cost (usually the shortest) by evaluating the expected cost from the starting point to the end point. However, although the A* algorithm performs well in static environments, its adaptability and efficiency in dynamically changing industrial environments are limited. Specifically, when temporary obstacles or unexpected movements of other AGVs appear in the environment, the A* algorithm needs to recalculate the path, which may cause significant delays and reduced efficiency in complex environments.

[0005] In addition, although existing heuristic path planning methods have made some improvements in improving search efficiency, such as by introducing more complex heuristic functions to estimate the cost from the current node to the target node, these methods often do not fully consider the specific constraints and requirements in actual industrial environments. For example, parking rules in specific areas, one-way restrictions, and obstacle avoidance points set to avoid conflicts are all factors that must be considered in practical applications. Path planning methods that ignore these factors may not be effectively implemented in real environments, thus affecting the overall performance and safety of the AGV system.

[0006] Furthermore, as industrial scenarios become increasingly complex and dynamic, the limitations of traditional and existing heuristic path planning methods become more obvious. For example, the task requirements on an industrial production line may change frequently, and material supply points and demand points may also be adjusted according to the production plan, which requires the AGV's path planning system to be able to respond to these changes in real time and quickly update the path planning. At the same time, in scenarios where multiple AGVs collaborate, how to effectively avoid conflicts and congestion between AGVs and ensure the smooth operation of the entire system has also become an important research and application challenge.

[0007] Therefore, developing an efficient path planning method that can fully adapt to complex dynamic environments and take into account specific traffic control measures and multi-AGV collaboration is of great significance for improving the operating efficiency and safety of AGV systems. This method not only needs to be able to handle dynamic changes in the environment, but also needs to be able to adjust path planning in real time to cope with changing task requirements, thereby ensuring that the AGV system can work stably and efficiently under various conditions. Summary of the invention

[0008] The present invention proposes an AGV path scheduling method based on the spatial A-star algorithm and traffic control. The method not only considers the shortest path, but also comprehensively considers the real-time changes in the dynamic environment and various traffic control measures to improve the efficiency and safety of path planning.

[0009] The technical solution adopted by the present invention is:

[0010] An AGV path scheduling method based on spatial A-star algorithm and traffic control, the method comprising:

[0011] (1) Introduction of spatial information: Collect current environmental data, identify and mark key spatial elements on the map; map the key spatial elements onto the environmental map; the key spatial elements include: obstacle avoidance points, traffic control areas, high-risk areas, one-way streets, and long two-way paths;

[0012] (2) Path planning: Set the starting point and the end point, use the improved spatial A-star algorithm to search the path in the environment, and use the key spatial elements described in step (1) as key information in the search process to participate in path planning;

[0013] Multiple traffic control rules are set in path planning to determine the travel path; the multiple traffic control rules include parking rules, one-way street restrictions, and obstacle avoidance point management;

[0014] (3) Dynamic monitoring: When the AGV executes the travel path, the environment is monitored in real time, and the path is replanned when the environment changes; and when the AGV executes the travel path, anomaly detection and collision prediction are performed, and the path of the AGV is adjusted in real time according to the results of anomaly detection and collision prediction;

[0015] (4) Continue the task: The AGV continues to perform the task according to the path updated in step (3) until it reaches the destination.

[0016] Furthermore, in step (1), the environment map is regularly updated based on real-time data to ensure the accuracy of the path planning information.

[0017] Furthermore, in step (2), the improved spatial A-star algorithm includes a heuristic factor ; represents the heuristic estimated cost from the current node to the target node;

[0018] Heuristic Factor The calculation formula is:

[0019] ;

[0020] From the current node The traditional estimated cost to the target node (using Euclidean distance and Manhattan distance to represent the traditional estimated cost);

[0021] It is an adjustment parameter used to balance the weight of traditional cost estimation and spatial information; Greater than 0;

[0022] Scoring spatial information for evaluating nodes spatial impact.

[0023] Furthermore, the spatial information score Based on the node to calculate the special spatial elements encountered when The calculation methods include:

[0024] 1) If the node If you are located in a high-risk area:

[0025] ;

[0026] in, is the preset risk value, , is a constant greater than 0; when the node If you are not in a high-risk area, ;

[0027] 2) If there is a detour path, and the cost of the detour path is lower than the original path cost, use the detour path cost advantage To adjust :

[0028] ;

[0029] in, is a negative value;

[0030] 3) For one-way streets and parking areas, The adjustment depends on the node Whether the restrictions of the corresponding area are met:

[0031] If the node Violation of one-way restrictions Increase a value;

[0032] If the node In a parking area, if parking is not allowed in the current parking area, then Increase a value;

[0033] It is expressed as:

[0034] ;

[0035] in, The value to increase is a positive number.

[0036] Furthermore, in step (2), the method of using the improved spatial A-star algorithm to search for paths in the environment specifically includes the following steps:

[0037] (2.1) Initialization: Open list Contains the starting node, sets the initial , ; Closed list is empty; Represents the traditional estimated cost from the current initial node start to the target node;

[0038] (2.2) Loop processing: select from the open list The smallest node As the current node; among them, represents the actual cost from the starting point to the current node, Heuristically estimate the cost from the current node to the target node; is the total cost from the starting point to the target node, including the actual cost and the heuristic estimated cost;

[0039] if If it is the target node, the path is found and the loop is exited; otherwise, Move into the closed list;

[0040] (2.3) Node expansion: For the current node Each adjacent node of :

[0041] calculate ,in, From the starting point to the node The actual cost, Represents the distance from the starting point to the current node The actual cost, Indicates the current node To Node the actual cost of

[0042] calculate ;in, Represents a slave node The heuristic estimated cost to the target node, Represents a slave node The traditional estimated cost to the target node, For Node The corresponding spatial information score;

[0043] calculate ;

[0044] if Not in the open list, or Less than before value, then update And set the node The parent node is ;

[0045] (2.4) Termination condition: If the open list is empty, it means that no feasible path has been found and the path search is terminated; otherwise, the found path is returned.

[0046] Furthermore, in step (2), the parking rules are used to ensure that the AGV can quickly find a parking location when it needs to park, and to avoid path blockage or interference with the movement of other AGVs caused by parking;

[0047] The one-way restriction is used to ensure that the AGV automatically recognizes and complies with the one-way restriction when planning a path, so as to improve the traffic efficiency in narrow areas;

[0048] The obstacle avoidance point management is implemented through fixed obstacle avoidance point calibration and dynamic obstacle avoidance point detection. When performing path planning, the obstacle avoidance points can be bypassed efficiently and safely through detours, speed adjustment, and parking waiting.

[0049] Furthermore, in step (3), the environment is monitored in real time, and when the environment changes, the method of replanning the path includes:

[0050] Environmental monitoring: Install a variety of sensors on the AGV to collect environmental data; use synchronous positioning and map building technology to process environmental data and generate real-time maps;

[0051] Information update: compare the current environment map with historical data, detect newly appeared obstacle avoidance points and traffic control areas, mark and record the changed areas in real time, and update the corresponding spatial information data; transmit the updated spatial information data to the central control system, which synchronizes the updated spatial information data to all running AGVs;

[0052] Path replanning: When environmental changes are detected, path replanning is triggered, or when there are no major environmental changes, the current path is re-evaluated regularly to obtain the optimal path.

[0053] Furthermore, in step (3), anomaly detection and collision prediction are performed when the AGV executes the travel path, specifically:

[0054] The abnormal detection includes obstacle approach detection and traffic rule violation detection; the obstacle approach detection uses laser radar and ultrasonic sensors to monitor obstacles around the AGV in real time, evaluate whether there is a potential collision risk, predict the intersection point with the AGV path according to the speed and trajectory of the obstacle, and issue an early warning; the traffic rule violation detection ensures compliance with current traffic control rules by monitoring the driving status of the AGV in real time;

[0055] The collision prediction uses a kinematic model to predict the future position of the AGV and its surrounding objects, estimates the motion trajectory based on speed, direction, and acceleration parameters, and predicts possible collision points; in a multi-AGV collaboration scenario, comprehensively considers the motion status of all AGVs, predicts possible path intersections, and plans avoidance strategies in advance.

[0056] Furthermore, in step (3), the path of the AGV is adjusted in real time according to the results of anomaly detection and collision prediction, and the path adjustment includes a real-time path adjustment mechanism and an emergency stop mechanism;

[0057] The path adjustment strategy of the real-time path adjustment mechanism is: adjusting the travel path of the AGV according to the collision prediction result;

[0058] If an unavoidable collision is predicted, the emergency stop mechanism is adopted, and the AGV immediately performs an emergency stop and notifies the control center; after the emergency stop, the AGV automatically enters the safety mode and waits for new path planning instructions. After the emergency is resolved, the path is replanned according to the updated environmental information and the unfinished task is continued.

[0059] Beneficial technical effects of the present invention:

[0060] Efficiency: The AGV path scheduling method provided by the present invention adopts an improved spatial A* algorithm (i.e., improved spatial A-star algorithm), which significantly optimizes the path search process by introducing spatial information as an additional heuristic factor; it can more accurately identify and avoid potential obstacles and inefficient areas, such as common congestion points or special restricted areas, thereby effectively reducing the size of the search space and the time for path calculation. This improvement enables the AGV to quickly find the optimal path when performing path planning, greatly improving the efficiency of path planning in complex environments, and enabling the AGV to respond and adjust its route in a short time to adapt to changes in real-time task requirements.

[0061] Safety: Safety is another major advantage of the present invention; by comprehensively considering various traffic control measures, such as parking rules, one-way restrictions, and dynamic management of obstacle avoidance points, this method can effectively avoid potential conflicts and dangers during the movement of AGVs. In addition, this method also takes into account the interaction between AGVs, and accurately controls the speed and interval of AGVs in specific areas to avoid possible mutual collisions or blockages, thereby ensuring the high safety of the entire AGV system. This comprehensive safety design not only ensures the safe driving of the AGV itself, but also provides safety protection for other equipment and personnel in the AGV operation area.

[0062] Adaptability: The dynamic adjustment mechanism is one of the core features of the present invention, which allows the AGV system to adjust the planned path in time when facing real-time environmental changes and other AGV status changes. This mechanism is based on complex real-time data analysis and decision-making algorithms, which can quickly identify new changes in the environment, such as temporary obstacles, changes in the work area, or other unexpected behaviors of AGVs, and re-plan the path accordingly. This capability enables the AGV system to flexibly respond to various uncertainties and dynamic changes in the working environment, maintain efficient operation, and minimize task delays or stagnation caused by environmental changes or other unexpected factors. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] On the contrary, the present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention as defined by the claims. Further, in order to make the public have a better understanding of the present invention, some specific details are described in detail in the detailed description of the present invention below. Those skilled in the art can fully understand the present invention without the description of these details.

[0065] Embodiment 1: An AGV path scheduling method based on spatial A-star algorithm and traffic control; the method comprises:

[0066] (1) Introduction of spatial information: Collect current environmental data, identify and mark key spatial elements on the map; and map the key spatial elements onto the environmental map;

[0067] The key spatial elements include: obstacle avoidance points, traffic control areas, high-risk areas, one-way streets, and long two-way paths; the traffic control areas include one-way streets, speed limit areas, and temporarily closed areas; among them, the high-risk areas refer to areas around dangerous equipment, suspended areas on the site, crowded areas, and other areas that may threaten people or the AGV itself;

[0068] (2) Path planning: Set the starting point and the end point, use the improved spatial A-star algorithm to search the path in the environment, and use the key spatial elements described in step (1) as key information in the search process to participate in path planning;

[0069] Multiple traffic control rules are set in path planning to determine the travel path; the multiple traffic control rules include parking rules, one-way restrictions, and obstacle avoidance point management; specifically, parking rules refer to considering preset parking rules in path planning and selecting appropriate parking points. Complying with one-way restrictions refers to identifying one-way areas and planning paths to ensure that traffic rules are not violated. Obstacle avoidance point management refers to identifying fixed and temporary obstacle avoidance points in the path and planning detours or obstacle avoidance strategies; the above multiple traffic control strategies are combined to make the final travel path decision.

[0070] (3) Dynamic monitoring: When the AGV executes the travel path, the environment is monitored in real time, and the path is replanned when the environment changes; and when the AGV executes the travel path, anomaly detection and collision prediction are performed, and the path of the AGV is adjusted in real time according to the results of anomaly detection and collision prediction;

[0071] (4) Continue the task: The AGV continues to perform the task according to the path updated in step (3) until it reaches the destination.

[0072] Step (1) of this embodiment includes:

[0073] Environmental preprocessing: The system first scans the environment and identifies all key spatial elements, including obstacle avoidance points, traffic control areas (such as one-way streets, speed limit areas, temporarily closed areas, etc.) and other special areas that may affect AGV path planning (such as high-risk areas around dangerous equipment).

[0074] Spatial element marking: The identified spatial elements are marked according to their attributes and functions and mapped to the environment map. These marking information will serve as an important reference in the path search process and participate in path planning as key information in the search process. In this way, the algorithm can prioritize or accelerate the passage of special areas where these key spatial elements are located when searching for paths, effectively avoiding unnecessary searches and potential path conflicts.

[0075] In step (1), the environment map is regularly updated based on real-time data to reflect changes in the environment and ensure the accuracy of the path planning information.

[0076] Specifically, the method for updating the environment map includes:

[0077] Real-time monitoring and data collection: Receive updated data from the environmental monitoring system regularly or as needed, including the appearance of new obstacles, changes in traffic control rules, etc.

[0078] Update process: parse the update data, update the relevant nodes in the environment map, modify their obstacle status, traffic control information, etc.

[0079] If necessary, rebuild the graphical connections within the area of ​​influence.

[0080] In step (2) of this embodiment, the improved spatial A-star algorithm includes a heuristic factor ; represents the heuristic estimated cost from the current node to the target node;

[0081] Heuristic Factor The calculation formula is:

[0082] ;

[0083] From the current node The traditional estimated cost to the target node (using Euclidean distance and Manhattan distance to represent the traditional estimated cost);

[0084] It is an adjustment parameter used to balance the weight of traditional cost estimation and spatial information; The value of should be tuned according to the actual application scenario to properly balance the impact of traditional estimated cost and spatial information on path planning; Greater than 0.

[0085] Scoring spatial information for evaluating nodes spatial impact;

[0086] Among them, the spatial information score Based on the node The special spatial elements encountered during calculation are used to calculate The calculation methods include:

[0087] 1) If the node If you are located in a high-risk area:

[0088] ;

[0089] in, is the preset risk value, , is a constant greater than 0; when the node If you are not in a high-risk area, ;in, The algorithm can be optimized through simulation experiments or historical data to improve its practicality and accuracy;

[0090] 2) If there is a detour path, and the cost of the detour path is lower than the original path cost, use the detour path cost advantage To adjust :

[0091] ;

[0092] in, is negative; that is, if there is a detour path and the cost of that path is relatively low (avoiding high-risk areas or high-congestion areas), then Should be reduced.

[0093] 3) For one-way streets and parking areas, The adjustment depends on the node Whether the restrictions of the corresponding area are met:

[0094] If a node violates a one-way street restriction, then Increase a value;

[0095] If the node In a parking area, if parking is not allowed in the current parking area, then Increase a value;

[0096] It is expressed as:

[0097] ;

[0098] in, The value to increase is a positive number.

[0099] Improved spatial A-star algorithm in designing heuristic factors In addition to considering the estimated cost (such as distance or time) from the current node to the target node, the algorithm also takes into account the impact of spatial information. This means that when evaluating the potential value of a node, the algorithm will comprehensively consider the special spatial elements that may be encountered when passing through the node, such as whether it will pass through a high-risk area and whether there is a better detour path. This comprehensive evaluation method makes path planning more accurate and efficient, especially in the face of complex and changing environments, which can significantly reduce search time and improve the quality of path planning.

[0100] In step (2) of this embodiment, the path planning includes:

[0101] Set the starting point and end point: the user defines or the system generates the current position of the AGV as the starting point and the target position as the end point;

[0102] Heuristic Function Design: Designing Heuristic Functions , considering the distance to the target point and the possibility and cost of passing through special areas.

[0103] Search algorithm implementation: Use the improved A-star algorithm to loop through the nodes in the open list and select the node with the lowest The node with the value is taken as the next step.

[0104] represents the actual cost from the starting point to the current node, Heuristically estimate the cost from the current node to the target node.

[0105] Path reconstruction: Starting from the end node, trace the parent node in reverse until you return to the starting point. This process constitutes the planned path.

[0106] In step (2), the method of using the improved spatial A-star algorithm to search for paths in the environment specifically includes the following steps:

[0107] (2.1) Initialization: Open list Contains the starting node, sets the initial , ; Closed list is empty; Represents the traditional estimated cost from the current initial node start to the target node;

[0108] (2.2) Loop processing: select from the open list The smallest node As the current node; among them, represents the actual cost from the starting point to the current node, Heuristically estimate the cost from the current node to the target node; is the total cost from the starting point to the target node, including the actual cost and the heuristic estimated cost;

[0109] if If it is the target node, the path is found and the loop is exited; otherwise, Move into the closed list;

[0110] (2.3) Node expansion: For the current node Each adjacent node of :

[0111] calculate ,in, From the starting point to the node The actual cost, Represents the distance from the starting point to the current node The actual cost, Indicates the current node To Node the actual cost of

[0112] calculate ;in, Represents a slave node The heuristic estimated cost to the target node, Represents a slave node The traditional estimated cost to the target node, For Node The corresponding spatial information score;

[0113] calculate ;

[0114] if Not in the open list, or Less than before value, then update And set the node The parent node is ;

[0115] (2.4) Termination condition: If the open list is empty, it means that no feasible path has been found and the path search is terminated; otherwise, the found path is returned.

[0116] By using the improved spatial A-star algorithm to search for paths in the environment, AGV can more intelligently respond to path planning challenges in complex dynamic environments and improve the efficiency and safety of the overall system.

[0117] In step (2) of this embodiment, the parking rule is used to ensure that the AGV can quickly find a suitable parking location when it needs to park, and avoid path blockage or interference with the movement of other AGVs caused by parking;

[0118] The one-way restriction is used to ensure that the AGV automatically recognizes and complies with the one-way restriction when planning a path, so as to improve the traffic efficiency in narrow areas;

[0119] The obstacle avoidance point management is implemented through fixed obstacle avoidance point calibration and dynamic obstacle avoidance point detection. When performing path planning, the obstacle avoidance points can be bypassed efficiently and safely through detours, speed adjustment, and parking waiting.

[0120] The parking rules specifically include:

[0121] 1) Parking spot selection: including real-time environmental assessment, priority sorting and dynamic selection mechanism;

[0122] Real-time environmental assessment: When performing tasks, the AGV obtains the current environment in real time, including the current location, target location, and surrounding parking locations; uses the sensor system and environmental map to identify the nearest available parking spot and assess the safety and availability of the available parking spot (the safety and availability assessment specifically includes whether the available parking spot has enough space and complies with traffic regulations);

[0123] Prioritization: Prioritize available parking spots based on their distance, environmental safety, and mission urgency. If multiple parking spots meet the criteria, select the closest spot that will not block other AGV paths.

[0124] Dynamic selection mechanism: If the AGV detects that a parking spot is occupied or a new available parking spot appears during the journey, the parking selection is adjusted in real time;

[0125] 2) Parking time limit:

[0126] Task priority considerations: Set parking time limits based on the nature of the task (e.g. material transportation, emergency tasks); shorter parking times may be allowed for high-priority tasks, while longer parking times may be allowed for low-priority tasks;

[0127] Dynamically adjust parking time: During the parking process, the AGV system continuously monitors the progress of the task and environmental changes. If the urgency of the task increases or the environment changes (such as new obstacles or traffic control information), the parking time is dynamically adjusted;

[0128] Notification and recovery mechanism: Before each parking, the estimated parking time is sent to the central control system for record; if the AGV needs to extend the parking time, it should immediately notify the control center and decide whether to continue waiting or resume the task according to the new task instructions;

[0129] 3) Emergency parking procedures include:

[0130] Automated emergency stop decision: When an emergency situation is detected (such as sudden obstacles, system failures, or other factors that endanger safety), the AGV system automatically decides to perform an emergency stop; the system calculates the safety of the surrounding environment before stopping and selects a safe location for parking;

[0131] Multi-level response mechanism: Depending on the severity of the emergency, the system responds in different levels; for lower-level emergencies, the AGV can slow down or adjust the path; for high-risk situations, the AGV should stop immediately and enter safety mode;

[0132] Task recovery and path replanning: After the emergency is resolved, the system automatically replans the path and resumes task execution in the shortest possible time; when replanning the path, priority will be given to avoiding areas where problems occurred previously.

[0133] In the present invention, the one-way street restriction specifically includes:

[0134] 1) Loading one-way street information, including:

[0135] Map initialization loading: Before path planning begins, the AGV loads relevant information of all one-way streets in the current environment from the central control system; the relevant information includes the starting point, end point and direction of the one-way street;

[0136] Dynamic update of one-way street information: real-time update of one-way street information (e.g., when encountering construction, temporary traffic control, etc., the direction of the one-way street may change), and the AGV receives and applies the latest information in a timely manner;

[0137] 2) One-way street check during path planning, including:

[0138] Path compliance check: When planning a path, the AGV system checks whether the planned path passes through a one-way street area and confirms whether the AGV's direction of travel complies with the direction restrictions of the one-way street. If the planned path does not comply with the direction of the one-way street, the path will be automatically replanned to ensure compliance with traffic regulations.

[0139] Priority path selection: In one-way areas, the path that matches the direction of travel is given priority to avoid unnecessary path replanning and reduce the complexity of path planning.

[0140] 3) Alternative route planning, including:

[0141] Automatically search for alternative routes: If the originally planned route is not feasible due to a one-way street restriction, the AGV system will automatically search for an alternative route; the alternative route should avoid the one-way street restriction as much as possible to ensure the continuity and efficiency of the route.

[0142] The goal of the obstacle avoidance point management is to provide the AGV with a flexible obstacle avoidance strategy, including detour, speed adjustment, and parking waiting, to ensure efficient and safe detour of the obstacle avoidance point. The obstacle avoidance point management specifically includes:

[0143] 1) Obstacle avoidance point calibration and management, including:

[0144] Fixed obstacle avoidance point calibration: In environmental preprocessing, known fixed obstacles (such as equipment and walls) are calibrated as obstacle avoidance points, and the fixed obstacle positions and impact ranges are marked on the map;

[0145] Dynamic obstacle avoidance point detection: Use the sensor system to detect dynamic obstacles in the environment (such as other AGVs, pedestrians, temporary objects, etc.) in real time, and mark the dynamic obstacles as temporary obstacle avoidance points; the information of temporary obstacle avoidance points is updated in real time to continuously track the movement of dynamic obstacles.

[0146] 2) Obstacle avoidance strategy selection, including:

[0147] Obstacle avoidance path planning: When planning the path, the impact of obstacle avoidance points is evaluated and a detour path is planned. The detour path should be as short as possible and avoid entering new obstacle avoidance areas.

[0148] Speed ​​adjustment strategy: If the detour path is not feasible or the detour distance is too long, the AGV system will choose to slow down and pass through the obstacle avoidance area; at this time, the AGV will pass at a safe speed and ensure that the obstacle avoidance point will not pose a threat to driving safety;

[0149] Stop and wait strategy: In the obstacle avoidance area, if the obstacle ahead cannot be bypassed or needs to be cleared, instruct the AGV to stop briefly until the obstacle avoidance area is open to traffic; during the stop and wait period, continuously monitor the obstacle status and be ready to resume path planning at any time;

[0150] 3) Obstacle avoidance behavior execution, including:

[0151] Automatically execute obstacle avoidance behavior: According to the choice of obstacle avoidance strategy, the AGV system automatically executes operations such as detour, deceleration or parking; when executing obstacle avoidance behavior, the path planning is adjusted in real time to ensure that the route after obstacle avoidance still meets the task requirements;

[0152] Dynamic obstacle avoidance adjustment: During the obstacle avoidance process, the changes of obstacles are monitored in real time. If the obstacle moves or disappears, the obstacle avoidance behavior will be dynamically adjusted, such as restoring normal speed or replanning the detour path.

[0153] The present invention sets a series of detailed parking rules, including but not limited to the selection of parking points, the limit of parking time, and the emergency parking processing mechanism under specific circumstances. Through these rules, the AGV can quickly find a suitable parking position when needed, while avoiding path blockage or other AGV interference caused by parking. In order to further improve the operating efficiency of the system and avoid conflicts between AGVs, the present invention strictly considers the one-way street restrictions in path planning. When planning its path, the AGV will automatically identify and comply with the one-way street restrictions to ensure that its direction of travel is consistent with the preset one-way street direction. This not only reduces the complexity of path planning, but also significantly improves the traffic efficiency of AGV in narrow areas. In a complex working environment, the existence of obstacle avoidance points is an important factor that must be considered in AGV path planning. The present invention provides AGV with detailed obstacle avoidance strategies, including detour obstacle avoidance, speed adjustment, and parking waiting, by accurately calibrating and managing obstacle avoidance points. These strategies enable AGV to flexibly adjust its path or behavior when encountering fixed or temporary obstacle avoidance points, ensuring that neither traffic control rules are violated nor operating efficiency is reduced. Through the above detailed multiple traffic control strategies, the AGV system can efficiently and safely perform tasks in complex industrial environments while avoiding potential problems caused by traffic rules or obstacles. The detailed description of these strategies ensures the robustness and adaptability of the system and improves the overall operation efficiency of the AGV.

[0154] In step (3) of this embodiment, the environment is monitored in real time. When the environment changes, the method of replanning the path includes:

[0155] Environmental monitoring: Install a variety of sensors on the AGV to collect environmental data; use synchronous positioning and map building technology to process environmental data and generate real-time maps;

[0156] Information update: compare the current environment map with historical data, detect newly appeared obstacle avoidance points and traffic control areas, mark and record the changed areas in real time, and update the corresponding spatial information data; transmit the updated spatial information data to the central control system, which synchronizes the updated spatial information data to all running AGVs;

[0157] Path replanning: When environmental changes are detected, path replanning is triggered, or when there are no major environmental changes, the current path is re-evaluated regularly to obtain the optimal path.

[0158] The purpose of environmental monitoring is to perceive changes in the environment, including the emergence of new obstacles, adjustments to traffic control rules, and changes in other dynamic factors. Specifically, the environmental monitoring method includes:

[0159] 1) Sensor layout and data collection:

[0160] Multi-sensor fusion: Multiple sensors are installed on the AGV, including LiDAR, cameras, ultrasonic sensors, and infrared sensors, to ensure all-round perception of the environment.

[0161] LiDAR is used to accurately measure the distance and shape of surrounding objects.

[0162] The camera provides visual information and recognizes road signs, traffic signs, etc.

[0163] Ultrasonic and infrared sensors are used to detect obstacles at close range.

[0164] Real-time data collection: The sensor system continuously collects environmental data at a high frequency to generate a real-time updated environmental map.

[0165] Use edge computing devices for preliminary data processing to reduce data transmission and processing delays.

[0166] 2) Data processing and environmental change detection:

[0167] Environmental change detection algorithm: Use SLAM (Simultaneous Localization and Mapping) technology to process environmental data and generate real-time maps. This is conventional positioning technology, such as UWB.

[0168] Compare current environment maps with historical data to detect new obstacle avoidance points, traffic control areas, or other important environmental changes.

[0169] Mark and record changed areas in real time to provide a basis for subsequent steps.

[0170] The purpose of information update is to ensure that the spatial information and traffic control rules stored in the system are always consistent with the actual environment. The information update includes:

[0171] 1) Information update mechanism:

[0172] Real-time update strategy: When new obstacle avoidance points or traffic control information are detected, the system immediately updates the corresponding spatial information database. The updated information includes the location and nature of obstacles, changes in traffic rules (such as speed limit zones, changes in the direction of one-way streets), etc.

[0173] Central control system synchronization: The updated information is transmitted to the central control system, which is responsible for synchronizing these updates to all running AGVs to ensure that the environmental information used by each AGV is the latest. The system supports different levels of update strategies, such as immediate global synchronization of high-priority changes and batch synchronization of low-priority changes in appropriate time windows.

[0174] The goal of path replanning is to recalculate the optimal path in a timely manner after detecting environmental changes to ensure that the AGV can continue to perform tasks efficiently and safely. The path replanning includes:

[0175] 1) Re-planning trigger conditions:

[0176] Environmental change trigger: When the system detects a significant environmental change (such as a new obstacle appearing on the planned path), it immediately triggers path replanning.

[0177] Periodic re-evaluation: In the absence of significant environmental changes, the system periodically re-evaluates the current path to ensure it is still the optimal path.

[0178] 2) Application of path planning algorithm:

[0179] Use the improved spatial A-star algorithm (see step (2) of the present invention, path planning step): recalculate the optimal path from the current node to the target node, taking into account the latest spatial information (including new obstacle avoidance points and updated traffic control areas). According to the dynamic information of the current environment, adjust the spatial information score S(n) in the heuristic factor to optimize the path selection.

[0180] Path Alternatives: While calculating the main path, calculate alternative paths to cope with possible further environmental changes. If the main path becomes infeasible during execution, you can quickly switch to the alternative path, reducing the time and resource consumption of path adjustment.

[0181] In step (3) of this embodiment, anomaly detection and collision prediction are performed when the AGV executes the travel path, specifically:

[0182] The goal of anomaly detection is to continuously monitor abnormal situations that may lead to collisions or other safety issues during path execution.

[0183] The abnormal detection includes obstacle approach detection and traffic rule violation detection; the obstacle approach detection uses laser radar and ultrasonic sensors to monitor obstacles around the AGV in real time, evaluate whether there is a potential collision risk, predict the intersection with the AGV path according to the speed and trajectory of the obstacle, and issue an early warning; the traffic rule violation detection is to monitor the driving status of the AGV in real time to ensure compliance with current traffic control rules (such as speed limits, one-way rules, etc.), and immediately trigger corresponding corrective measures (such as deceleration and parking) if it is detected that the AGV is about to violate traffic rules;

[0184] Abnormal data processing, including abnormal data classification: categorize detected abnormal situations into different levels (such as low risk, high risk) and take corresponding treatment measures. Low-risk abnormalities may only require fine-tuning of the route, while high-risk abnormalities may require immediate re-routing or emergency stopping.

[0185] Real-time feedback mechanism for abnormal data: The abnormal detection system is linked with the central control system to feed back abnormal information to the control center for more comprehensive analysis and processing. If abnormal situations occur frequently, the control center can dispatch more resources or modify the current AGV task plan.

[0186] The goal of collision prediction is to predict potential collision events and adjust the route in advance to avoid actual collisions.

[0187] The collision prediction uses a kinematic model to predict the future position of the AGV and its surrounding objects, estimates the motion trajectory based on speed, direction, and acceleration parameters, and predicts possible collision points; in a multi-AGV collaboration scenario, comprehensively considers the motion status of all AGVs, predicts possible path intersections, and plans avoidance strategies in advance.

[0188] Specifically, prediction based on motion models: Use kinematic models to predict the future positions of the AGV and its surrounding objects, and estimate their motion trajectories based on parameters such as speed, direction, acceleration, etc. The model can consider the interaction between the AGV and other dynamic objects (such as other AGVs or pedestrians) and predict possible collision points.

[0189] Multi-AGV collaborative collision prediction: In a multi-AGV collaborative scenario, the system will comprehensively consider the motion status of all AGVs, predict possible path intersections, and plan avoidance strategies in advance.

[0190] Optimize prediction accuracy and dynamically adjust prediction window: According to the speed of AGV and the complexity of the environment, dynamically adjust the prediction time window and spatial range to improve prediction accuracy. In high-speed or complex environments, increase prediction frequency and accuracy; in low-speed or simple environments, reduce the computational burden.

[0191] In step (3) of this embodiment, the path of the AGV is adjusted in real time according to the results of anomaly detection and collision prediction, and the path adjustment includes a real-time path adjustment mechanism and an emergency stop mechanism;

[0192] The path adjustment strategy of the real-time path adjustment mechanism is: adjusting the travel path of the AGV according to the collision prediction result;

[0193] If an unavoidable collision is predicted, the emergency stop mechanism is adopted, and the AGV immediately performs an emergency stop and notifies the control center; after the emergency stop, the AGV automatically enters the safety mode and waits for new path planning instructions. After the emergency is resolved, the path is replanned according to the updated environmental information and the unfinished task is continued.

[0194] Specifically, the goal of the path adjustment in the present invention is to adjust the path of the AGV in real time according to the results of anomaly detection and collision prediction to avoid potential safety risks.

[0195] The real-time path adjustment mechanism includes:

[0196] Path adjustment strategy: According to the collision prediction results, the AGV's path is adjusted, which can be fine-tuning (such as avoiding the obstacle area it is about to enter) or re-planning the path.

[0197] In high-risk situations, the system can prioritize a safer path, even if it may not be the shortest path.

[0198] Speed ​​adjustment and priority setting: In the process of dynamically adjusting the path, the system can increase the reaction time and reduce the risk of collision by reducing the AGV speed. In a multi-AGV scenario, the system dynamically adjusts the priority of the AGV according to the importance and urgency of the task to ensure the priority execution of critical tasks.

[0199] Path adjustment in emergency situations with an emergency stop mechanism: If an unavoidable collision or other major safety risks are predicted, the system can immediately perform an emergency stop and notify the control center.

[0200] After an emergency stop, the system automatically enters safety mode and waits for new path planning instructions.

[0201] Path recovery and task continuation: After the emergency is resolved, the system replans the path based on the updated environmental information and continues to execute the unfinished task. If the task is interrupted, the system will re-evaluate the current task priority and adjust the task execution order.

[0202] Through the above detailed scheme, the improved spatial A-star algorithm can dynamically adjust the path in a complex industrial environment and update the spatial information in real time, ensuring that the AGV can perform tasks efficiently and safely in any environment. This scheme combines advanced sensor technology, real-time data processing and intelligent path planning algorithms, greatly improving the adaptability and robustness of the AGV system.

[0203] The AGV path scheduling method includes the following steps: First, the AGV is assigned tasks according to the current task queue and vehicle status. Task assignment is based on multiple factors, including but not limited to the idle state of the AGV, the current position and the task priority. After the task is assigned, the system starts the path planning process, and the AGV calculates the optimal driving path from the starting point to the end point through the AGV path scheduling method provided by the present invention. The path planning takes into account obstacles, traffic conditions and other dynamic factors. In the task execution phase, the AGV moves according to the planned path, performs tasks and performs real-time monitoring to ensure accurate navigation and task completion. In order to avoid path conflicts between multiple AGVs, the system introduces a collision avoidance mechanism to ensure safe interaction between AGVs through spatial information planning and traffic control strategies. After the AGV completes the task, the system updates the AGV status to "idle" or enters the charging and maintenance state, and records relevant operation data to optimize the scheduling and path planning of subsequent tasks. The algorithm improves the overall efficiency and safety of the AGV system through dynamic task allocation, path planning and collision avoidance strategies.

[0204] The present invention provides an advanced AGV path planning method, which shows significant advantages over existing static planning methods and technologies based on deterministic search conditions (CSC). In traditional static path planning methods, once the path is planned and determined, the adaptability to environmental changes is poor, resulting in low efficiency in dynamically changing industrial environments. Although the scheduling method based on CSC is effective in deterministic environments, it often lacks sufficient adaptability and scalability when facing real-time changes and complex scenarios. In contrast, the path planning method provided by the present invention introduces an improved spatial A* algorithm, combined with dynamic monitoring and real-time adjustment mechanisms, so that the AGV can respond to environmental changes in real time during the execution of tasks and efficiently handle common problems such as temporary obstacles. This method not only improves the flexibility and response speed of path planning, but also significantly improves the safety and stability of path execution through precise parking rules and one-way restrictions. In addition, for the management of obstacle avoidance points in complex working scenarios, the present invention further ensures that the AGV will not violate traffic control rules or reduce operating efficiency while maintaining high efficiency through detailed obstacle avoidance strategies. In summary, compared with the existing technology, the technology of the present invention not only optimizes the overall efficiency and safety of path planning, but also significantly improves the adaptability and robustness in dynamic and uncertain environments, providing an efficient, safe and flexible solution for AGV scheduling in modern automated logistics and production systems.

[0205] Compared with the planning method based on learning, the technical solution of the present invention has obvious advantages in practicality, reliability and immediacy. The present invention provides a more immediate and deterministic path planning solution through an improved spatial A* algorithm. This method does not rely on large data sets or long training processes, but directly uses the spatial information of the environment to navigate efficiently, so that it can quickly adapt to real-time changes in the environment. This is especially important for industrial environments, because in such environments, AGVs may need to respond immediately to changes in production demand or temporary layout adjustments. By introducing dynamic monitoring and real-time adjustment mechanisms, AGVs can quickly respond to environmental changes and ensure stable operation while maintaining the efficiency and safety of path planning. In addition, the method of the present invention is also more advantageous than learning-based planning methods in terms of ensuring the stability and reliability of operations, because it avoids deviations and erroneous learning behaviors that may occur during the learning process. This is a consideration that cannot be ignored in the field of industrial automation that requires high safety.

[0206] In the path planning method of the present invention, the AGV takes special consideration of spatial information when planning its travel path to ensure that the optimization of the path is not only based on the principle of the shortest distance, but also fully considers the spatial layout and real-time changes of the environment. By introducing the improved spatial A-star algorithm, the method uses spatial information as an important heuristic factor to optimize the search process. This improvement means that when performing path search, the algorithm not only relies on traditional cost estimates (such as distance or time cost), but also considers special elements in the spatial layout, such as obstacle avoidance areas, high-risk areas, and areas affected by specific traffic control measures.

[0207] The present invention identifies key spatial elements that may affect the efficiency and safety of AGV travel, such as narrow passages, common congestion points, and sensitive areas that require special avoidance, by pre-analyzing and marking spatial information on the map. In the actual path planning process, the AGV will give priority to or avoid certain areas based on this spatial information, thereby ensuring the shortest path while avoiding possible safety hazards and inefficiency. In addition, this method also supports dynamic updating of spatial information to reflect immediate changes in the environment, such as temporary construction areas, to ensure that the AGV's path planning can adapt to environmental changes in real time, thereby improving the overall adaptability and flexibility of the AGV system.

[0208] By incorporating spatial information into path planning considerations, the present invention significantly improves the navigation efficiency and safety of AGV in complex environments. In particular, when faced with a dynamically changing operating environment, path planning and adjustment can be completed more accurately and quickly, significantly improving the operating efficiency and reliability of the AGV system.

[0209] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An AGV path scheduling method based on spatial A-star algorithm and traffic control, characterized in that: The method comprises: (1) Introduction of spatial information: Collect current environmental data, identify and mark key spatial elements on the map; and map the key spatial elements onto the environmental map; (2) Path planning: Set the starting point and the end point, use the improved spatial A-star algorithm to search the path in the environment, and use the key spatial elements described in step (1) as key information in the search process to participate in the path planning; set multiple traffic control rules in the path planning to determine the travel path; the multiple traffic control rules include parking rules, one-way street restrictions, and obstacle avoidance point management; (3) Dynamic monitoring: The AGV monitors the environment in real time while executing the path, and replans the path when encountering environmental changes; and performs anomaly detection and collision prediction while the AGV executes the path, and adjusts the path of the AGV in real time based on the results of anomaly detection and collision prediction; (4) Continue the task: The AGV continues to perform the task according to the path updated in step (3) until it reaches the destination; In step (2), the improved spatial A-star algorithm includes a heuristic factor h′(n); h′(n) represents the heuristic estimated cost from the current node to the target node; The calculation formula of the heuristic factor h′(n) is: h′(n)=h(n)+α*S(n); h(n) is the traditional estimated cost from the current node n to the target node; α is a tuning parameter used to balance the weight of traditional cost estimation and spatial information; S(n) is the spatial information score used to evaluate the spatial impact of node n; The spatial information score S(n) is calculated based on the special spatial elements encountered when passing through node n. The calculation method of S(n) includes: 1) If node n is located in a high-risk area, then: S(n)=S(n)+R(n); Wherein, R(n) is the preset risk value, R(n)=β, β is a constant greater than 0; when node n is not located in the high-risk area, R(n)=0; 2) If there is a detour path, and the cost of the detour path is lower than the original path cost, use the detour path cost advantage C 绕行 (n) to adjust S (n): S(n)=S(n)+C 绕行 (n); Among them, C 绕行 (n) is a negative value; 3) For one-way streets and parking areas, the adjustment of S(n) depends on whether node n meets the restrictions of the corresponding area: If node n violates the one-way restriction, S(n) increases by one value; If node n is in a parking area, and if parking is not allowed in the current parking area, S(n) increases by one value; S(n) is expressed as: S(n)=S(n)+T(n); Among them, T(n) is the increased value and is a positive number.

2. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (1), the environment map is regularly updated based on real-time data to ensure the accuracy of the path planning information.

3. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (2), the method of using the improved spatial A-star algorithm to search for paths in the environment specifically includes the following steps: (2.1) Initialization: The open list Open contains the start node, and the initial g(n) is set to 0, h′(n) = h(start) + α × S(start); the closed list Closed is empty; where h(start) represents the traditional estimated cost from the current start node start to the target node; (2.2) Loop processing: Select the node n with the smallest f(n) = g(n) + h′(n) from the open list. min as the current node; where g(n) represents the actual cost from the starting point to the current node, h′(n) is the heuristic estimated cost from the current node to the target node; f(n) is the total cost from the starting point to the target node, including the actual cost and the heuristic estimated cost; If n min is the target node, the path is found and the loop is exited; otherwise, n min Move into the closed list; (2.3) Node expansion: For the current node n min Each adjacent node n neighbor : Calculate g(n neighbor )=g(n min )+Cost(n min , n neighbor ), where g(n neighbor ) represents the distance from the starting point to node n neighbor The actual cost, g(n min ) represents the distance from the starting point to the current node n min The actual cost, Cost(n min , n neighbor ) represents the current node n min To node n neighbor the actual cost of Calculate h′(n neighbor )=h(n neighbor )+α*S(n neighbor ); where h′(n neighbor ) represents the slave node n neighbor The heuristic estimated cost to the target node, h(n neighbor ) represents the slave node n neighbor To the target node, S(n neighbor ) is node n neighbor The corresponding spatial information score; Calculate f(n neighbor )=g(n neighbor )+h′(n neighbor ); If n neighbor is not in the open list, or f(n neighbor ) is less than the previous f(n neighbor ) value, then update f(n neighbor ) and set node n neighbor The parent node is n min ; (2.4) Termination condition: If the open list is empty, it means that no feasible path has been found and the path search is terminated; otherwise, the found path is returned.

4. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (2), the parking rules are used to ensure that the AGV can quickly find a parking location when it needs to park, and to avoid path blockage or interference with the movement of other AGVs caused by parking; The one-way restriction is used to ensure that the AGV automatically recognizes and complies with the one-way restriction when planning a path, so as to improve the traffic efficiency in narrow areas; The obstacle avoidance point management is implemented through fixed obstacle avoidance point calibration and dynamic obstacle avoidance point detection. When performing path planning, the obstacle avoidance points can be bypassed efficiently and safely through detours, speed adjustment, and parking waiting.

5. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (3), the environment is monitored in real time. When the environment changes, the method of replanning the path includes: Environmental monitoring: Install a variety of sensors on the AGV to collect environmental data; use synchronous positioning and map building technology to process environmental data and generate real-time maps; Information update: compare the current environment map with historical data, detect newly appeared obstacle avoidance points and traffic control areas, mark and record the changed areas in real time, and update the corresponding spatial information data; transmit the updated spatial information data to the central control system, which synchronizes the updated spatial information data to all running AGVs; Path replanning: When environmental changes are detected, path replanning is triggered, or when there are no major environmental changes, the current path is re-evaluated regularly to obtain the optimal path.

6. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (3), anomaly detection and collision prediction are performed when the AGV executes the travel path, specifically: The abnormal detection includes obstacle approach detection and traffic rule violation detection; the obstacle approach detection uses laser radar and ultrasonic sensors to monitor obstacles around the AGV in real time, evaluate whether there is a potential collision risk, predict the intersection point with the AGV path according to the speed and trajectory of the obstacle, and issue an early warning; the traffic rule violation detection ensures compliance with current traffic control rules by monitoring the driving status of the AGV in real time; The collision prediction uses a kinematic model to predict the future position of the AGV and its surrounding objects, estimates the motion trajectory based on speed, direction, and acceleration parameters, and predicts possible collision points; in a multi-AGV collaboration scenario, comprehensively considers the motion status of all AGVs, predicts possible path intersections, and plans avoidance strategies in advance.

7. The AGV path scheduling method based on the spatial A-star algorithm and traffic control according to claim 1 is characterized in that: In step (3), the path of the AGV is adjusted in real time according to the results of anomaly detection and collision prediction, and the path adjustment includes a real-time path adjustment mechanism and an emergency stop mechanism; The path adjustment strategy of the real-time path adjustment mechanism is: adjusting the travel path of the AGV according to the collision prediction result; If an unavoidable collision is predicted, the emergency stop mechanism is adopted, and the AGV immediately performs an emergency stop and notifies the control center; after the emergency stop, the AGV automatically enters the safety mode and waits for new path planning instructions. After the emergency is resolved, the path is replanned according to the updated environmental information and the unfinished task is continued.