AGV obstacle avoidance path planning method and system for intelligent storage

Abstract

The invention provides an AGV obstacle avoidance path planning method and system for intelligent storage, and relates to the technical field of path planning. Comprising the steps of collecting warehousing data; constructing a map reflecting the topological structure of the storage space by adopting a topological map; obtaining an initial path by using a pruning optimized Dijkstra algorithm according to a set starting point and an end point; setting a repulsive force factor and a safe buffer area to optimize the initial path to obtain a secondary optimization path; and judging whether the environment is changed or not, and if so, recalculating a new obstacle avoidance path according to the secondary optimization path. According to the method, the initial path is constructed through the pruning optimized Dijkstra algorithm, the repulsive force factor and the safety buffer area are set to optimize the initial path, the overall efficiency and safety of warehouse cargo carrying are improved, and the adaptability of a warehouse system to complex environment changes is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of path planning technology, and in particular to an AGV obstacle avoidance path planning method for intelligent warehousing and an AGV obstacle avoidance path planning system for intelligent warehousing. Background Art

[0002] With the booming development of industries like e-commerce and manufacturing, warehousing and logistics business volumes have increased dramatically, placing higher demands on the efficiency, precision, and intelligence of warehousing operations. Traditional warehousing models, relying on manual handling and simple mechanical assistance, are no longer able to meet the demands of large-scale, high-frequency cargo handling. This has led to the emergence of smart warehousing, which aims to achieve efficient and precise control of warehousing processes through automated equipment and information technology. AGVs (automated guided vehicles), key automated handling equipment in smart warehousing, must operate stably in complex and dynamically changing warehouse environments. Effectively avoiding obstacles and rationally planning paths are key technical requirements for ensuring efficient operation, driving the continuous development of AGV obstacle avoidance and path planning technology.

[0003] There are different types of shelf layouts in storage facilities, such as traditional row shelves and high-rise shelves in automated warehouses. There are aisles of varying widths between the shelves, and there are often temporarily stacked goods and forklifts used for replenishment in the aisles. For AGVs to shuttle through such complex spatial structures, reliable obstacle avoidance path planning is required to avoid collisions and ensure smooth passage.

[0004] Warehousing operations are dynamic, with goods constantly being put into storage, sorted, and shipped out. During this process, the stacking location and space occupied by the goods are constantly changing, and new obstacles may appear at any time. The situation on the AGV's operating path is uncertain, so corresponding path planning technology is needed to respond to these changes in real time and ensure that it can flexibly adjust its route to avoid obstacles. Summary of the Invention

[0005] The present invention provides an AGV obstacle avoidance path planning method and system for intelligent warehousing, which are used to solve the defects of the prior art that it is difficult to minimize losses and is easy to be lost.

[0006] In one aspect, the present invention provides an AGV obstacle avoidance path planning method for intelligent warehousing, comprising: S1: Scan the storage environment to obtain storage data.

[0007] S2: Use topological maps to divide warehouse data into different grids and construct a topological structure map of the warehouse space.

[0008] S3: Use the Dijkstra algorithm with pruning optimization to generate an initial path based on the set starting and ending points on the warehouse space topology map.

[0009] S4: Set the repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path.

[0010] S5: Set a safety buffer zone so that the first optimized path can cope with the slight deviation during the AGV driving process according to the safety buffer zone and obtain the second optimized path.

[0011] S6: Monitor the AGV's operating status in real time to determine whether the environment has changed. If so, recalculate the new obstacle avoidance path based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environmental map.

[0012] According to an AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S1, the warehousing data includes obstacle position data, passable channel data, shelf position data, AGV starting point and target point data, and AGV real-time position data.

[0013] According to the AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S2, the specific steps of constructing the warehouse space topology map are: S21: Analyze the intersection nodes of aisles between shelves and the connection locations between different functional areas based on warehouse data.

[0014] S22: Determine whether there is a passable passage between the nodes based on the intersection nodes and the connection positions. If so, construct an edge between the two nodes and assign edge attribute characteristics to each edge.

[0015] S23: Integrate the nodes and edges to obtain a map framework of the spatial topological structure.

[0016] S24: Establish a dynamic update mechanism to update the map framework of the initial spatial topology structure in real time.

[0017] According to the AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S3, the specific steps of finding the path using the Dijkstra algorithm with pruning optimization are as follows: S31: Put the starting point into the visited node set, initialize a distance array, set the distance from the starting point to itself as the leader, initialize the distance from the starting point to all other nodes to infinity, and record the predecessor node of each node as empty.

[0018] S32: Traverse all adjacent nodes of the starting point, update the distance values ​​of the corresponding adjacent nodes in the distance array according to the weights of the edges from the starting point to the adjacent nodes, and set the predecessor nodes of these adjacent nodes as the starting point.

[0019] S33: Select the node with the smallest distance from the unvisited nodes and add it to the visited set.

[0020] S34: Repeat the step of adding the node to the visited set until the end point is added to the visited node set.

[0021] S35: When the end point is added to the visited node set, start from the end point, backtrack from the predecessor node, obtain each node from the starting point to the end point in sequence, and combine these nodes in order to obtain the initial path.

[0022] According to the AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S33, the specific steps of selecting a node to join the set include: S331: Select the node with the minimum current distance value from the nodes that have not been visited and add it to the visited node set.

[0023] S332: Recheck and update all adjacent nodes of the node with the minimum current distance value to determine whether the distance from all adjacent nodes to the starting point is smaller than the distance from the node with the minimum current distance value to the starting point. If so, update the distance value and add the predecessor node of the adjacent node to the node of the visited set.

[0024] According to the AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, the specific steps of path planning adjustment in step S4 are: S41: Define corresponding repulsion functions for different types of obstacles.

[0025] S42: Calculate the repulsive force vector generated by each obstacle according to the repulsive force function, superimpose the repulsive force vectors of all obstacles to obtain a total repulsive force vector, and synthesize the total repulsive force vector with the attraction vector to obtain the final resultant force direction of the AGV.

[0026] S43: As the AGV moves and the position of the obstacle changes, the distance parameters between each obstacle and the AGV are updated according to the repulsive force vector.

[0027] S44: recalculating the repulsion function value according to the distance parameter, and recalculating the resultant force direction according to the repulsion function value to obtain an optimized path.

[0028] According to an AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S42, the repulsion function formula is expressed as:

[0029] In the formula, k rep is the repulsion coefficient, d is the distance between AGV and obstacle i, and ρ is the radius of the obstacle’s influence range. is the vector pointing from the obstacle to the AGV, , the obstacle coordinates are (x obs ,y obs ), AGV coordinates are (x agv ,y agv ).

[0030] According to the AGV obstacle avoidance path planning method for intelligent warehousing provided by the present invention, in step S5, the specific steps for dealing with slight deviations during the AGV driving process are: S51: Determine the scope of the safety buffer zone based on the warehouse data.

[0031] S52: Compare the planned position of the primary optimized path with the range of the safety buffer zone to determine whether the AGV's driving has deviated. If so, adjust the primary optimized path according to the deviation and the range of the safety buffer zone to obtain a secondary optimized path.

[0032] According to the present invention, an AGV obstacle avoidance path planning method for intelligent warehousing also includes the collaborative operation of multiple AGVs, dynamically adjusting the path of each AGV through a conflict-based search algorithm, independently planning a preliminary path for each AGV, and determining whether there is a conflict between the paths through detection. If so, coordination and replanning are carried out.

[0033] On the other hand, the present invention also provides an AGV obstacle avoidance path planning system for intelligent warehousing, comprising: The data acquisition module is used to scan the warehouse data to obtain the warehouse data.

[0034] The data partitioning module is used to divide the warehouse data into different grids using a topological map and construct a map reflecting the topological structure of the warehouse space.

[0035] The path construction module is used to use the Dijkstra algorithm with pruning optimization to obtain the initial path based on the set starting and ending points on the constructed map.

[0036] The path optimization module is used to set the repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path.

[0037] The secondary optimization module is used to set a safety buffer zone so that the primary optimization path can cope with slight deviations during the AGV's driving process according to the safety buffer zone to obtain a secondary optimization path.

[0038] The real-time update module is used to monitor the AGV's operating status in real time to determine whether the environment has changed. If so, the new obstacle avoidance path is recalculated based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environmental map.

[0039] The present invention provides an AGV obstacle avoidance path planning system method and system for intelligent warehousing. By using the Dijkstra algorithm with pruning optimization to construct an initial path and setting a repulsion factor and a safety buffer zone to optimize the initial path, the beneficial effects achieved are as follows: A topological map method focuses on capturing the connections between locations in the environment and between them. By rationally identifying nodes, constructing edges, and assigning attributes, combined with a dynamic update mechanism, the resulting map accurately reflects the real-time state of the warehouse environment, providing realistic and timely foundational data for path planning, facilitating more accurate and optimal path planning. During path planning, an improved Dijkstra algorithm with pruning optimization is employed. This algorithm comprehensively considers obstacle avoidance and efficiency in the warehouse scenario to guide the search direction and quickly determine an initial path. By introducing a repulsive force factor and referencing the artificial potential field method, the path is automatically optimized to avoid obstacles based on a defined repulsive force function and resultant force calculation. A safety buffer is then established, and by monitoring minor deviations and fine-tuning the path, a secondary optimized path is obtained. This series of optimizations results in a safer, more efficient, and adaptable path to the complexities of the warehouse environment. During path planning, an improved Dijkstra algorithm with pruning optimization is employed to comprehensively consider obstacle avoidance and efficiency in the warehouse scenario to guide the search direction and quickly determine an initial path. By introducing repulsive forces and referring to the artificial potential field method, the path is optimized to automatically avoid obstacles based on a defined repulsive force function and resultant force calculation. A safety buffer zone is established, and by monitoring minor deviations and fine-tuning the path, a secondary optimized path is obtained. This series of optimizations makes the planned path safer, more efficient, and more adaptable to complex warehouse environments.

[0040] Real-time monitoring of AGV operating status and environmental changes. Once a change in the environment is detected, the system can quickly use the current AGV position as a new starting point, determine the target point as needed, recalculate the obstacle avoidance path based on the updated map, and perform rationality verification and correction. The accurate path information is transmitted to the control system to ensure that the AGV can travel along the new path in a timely manner. At the same time, the environmental monitoring and path adjustment process is continuously executed in a cycle, so that the AGV can operate continuously, safely and efficiently in the warehouse environment, effectively avoiding safety accidents such as collisions, greatly improving the overall efficiency and safety of warehouse cargo handling, and enhancing the warehouse system's adaptability to complex environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 This is a flow chart of an AGV obstacle avoidance path planning method for intelligent warehousing provided by an embodiment of the present invention; Figure 2 This is a module diagram of an AGV obstacle avoidance path planning system for intelligent warehousing provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0044] The following combination Figure 1-Figure 2 The present invention describes an AGV obstacle avoidance path planning method and system for intelligent warehousing.

[0045] Figure 1 This is a flow chart of an AGV obstacle avoidance path planning method for intelligent warehousing provided by an embodiment of the present invention.

[0046] like Figure 1 As shown, an embodiment of the present invention provides an AGV obstacle avoidance path planning method for intelligent warehousing, the method comprising: S1: Warehouse data includes obstacle location data, navigable aisle data, shelf location data, AGV starting and destination data, real-time AGV location data, other AGV location information, aisle availability data, and environmental map updates. Ultrasonic sensors can be used as a supplementary detection method at close range, especially for detecting obstacles in corners and low locations. Visual sensors identify different types of obstacles and obtain richer image feature information. These sensors are strategically installed at different locations on the AGV (automated guided vehicle) to ensure full, comprehensive coverage of the area surrounding the AGV, enabling real-time perception of the surrounding environment.

[0047] Obstacle location data includes: the coordinates of the obstacle. Static obstacles include shelves and columns, and dynamic obstacles include temporarily stacked goods, forklifts, and other AGVs. Obstacle location data is detected in real time using lidar, ultrasonic sensors, and visual sensors.

[0048] Obstacle types include fixed obstacles and mobile obstacles, which are used to define different repulsion functions. The radius of the obstacle's impact range determines the setting of the safety buffer zone.

[0049] Passable channel data includes: channel intersection node coordinates (as nodes in the topological map), channel width, slope, permitted AGV types (as edge attributes), and channel real-time passability status. This is obtained through sensor scanning combined with warehouse layout analysis.

[0050] Shelf location data includes shelf coordinates, including shelf corners, the connection between the storage and sorting areas, and shelf functional areas, including storage, sorting, and shipping areas. This data is used for cross-area route planning. Warehouse layout information is combined with sensor positioning.

[0051] AGV starting point and target point data include: starting point coordinates, AGV current position, target point coordinates, task instructions or warehouse management system allocation.

[0052] AGV real-time location data includes: AGV current coordinates and orientation, AGV driving speed, acceleration and other motion states.

[0053] Other AGV location information includes the real-time coordinates, driving direction, and speed of other AGVs, as well as the path planning information of each AGV. This information is obtained using the communication module in the multi-AGV collaborative system.

[0054] Aisle accessibility data includes: whether the aisle is blocked, including temporarily stacked goods and faulty AGVs. Aisle accessibility includes width, slope, and the number of temporary obstacles. Real-time monitoring using sensors combined with a dynamic update mechanism.

[0055] Environmental map update data includes: the location and type of new obstacles, the removal mark of disappeared obstacles, and the real-time location and mission status of other AGVs. Sensor fusion and dynamic map update algorithms are used.

[0056] Preprocess the collected warehouse data, such as removing noise data generated by the sensor itself, unifying the coordinates and synchronizing the time of data from different sensors to ensure that each data can accurately reflect the actual environmental conditions at the same time, and integrating the processed data into a unified format to facilitate subsequent analysis and judgment. Data preprocessing includes: Remove noise data caused by the sensor itself.

[0057] Unify the coordinates and synchronize the time of data from different sensors.

[0058] Integrate the processed data into a unified format.

[0059] S2: Using the information collected by the sensors, a topological map is used to mark the corresponding grid as an impassable area based on the location of obstacles detected by the sensors. At the same time, key information such as the passable aisles, shelf locations, AGV starting points and destination points are marked to construct a map that can reflect the real-time environmental status of the warehouse, providing basic data support for subsequent path planning. In addition, as the AGV moves and the environment changes, the map content is continuously updated to ensure the accuracy and timeliness of the map. The focus is on describing the various locations in the environment and the connection between locations. Through the use of sensors to perceive key locations in the environment, such as the intersection of aisles and the corners of shelves, as nodes, the passable paths between nodes are abstracted as edges, and the edges are accompanied by some attribute information, such as distance, difficulty of passage, etc.

[0060] The following are the specific steps for using the topological mapping method to construct a map reflecting the topological structure of the warehouse space in a warehouse environment: S21: The storage environment is scanned by sensors installed on the AGV to obtain environmental information, and the intersection positions of the aisles between the shelves and the connection positions between different functional areas are analyzed. These intersections are usually key decision points for path selection, and they are identified as nodes in the topological map. For example, in a warehouse with multiple rows of shelves and crisscrossing aisles, locations such as "cross" intersections and "T" intersections are marked as nodes. The connection positions between different functional areas are important nodes for AGVs to switch paths when performing different task stages. The entrance from the storage area to the sorting area, or the transition position from the sorting area to the shipping area, are all defined as nodes of the topological map to facilitate the subsequent planning of cross-regional transportation paths.

[0061] Columns, elevator entrances, fire-fighting facilities and other surrounding locations in the warehouse may affect the passage of AGVs or be places where AGVs need to bypass or dock. These locations are also considered as nodes, so that the topological map can more comprehensively cover key locations in the warehouse environment.

[0062] S22: Based on sensor information and the actual warehouse layout, determine whether there are accessible passages between nodes. If two nodes are directly connected and can be used by AGVs, an edge is created between them, indicating a direct path connection. Each edge is assigned attributes, including distance, difficulty, and permitted AGV types. These attributes facilitate more informed decisions during subsequent path planning. If a direct path from a particular intersection to a node at the junction of another area is detected, an edge is created between the two nodes. Distance is determined by actual measurement or by estimating the length of the passage between the two nodes based on map scale. Difficulty factors include passage width, slope, and the likelihood of temporary obstacles. For example, narrow passages are more difficult to navigate, so a higher difficulty coefficient can be set. Permitted AGV types include those with varying load capacities and sizes, and the AGVs that can pass through the path are indicated based on passage conditions.

[0063] S23: Integrate the determined nodes and the edges between them to form a preliminary map framework that reflects the topological structure of the warehouse space. Use a graph data structure to store this topological map, facilitating subsequent computer query, calculation, and path planning operations.

[0064] S24: Due to the dynamic nature of the warehouse environment, activities such as the stacking and handling of goods may change the accessibility of aisles. Therefore, a dynamic update mechanism should be established. Sensors should be used to continuously monitor environmental changes. When a new obstacle blocks the aisle corresponding to an edge, or a previously inaccessible area becomes accessible, the status of the edges in the topological map should be updated promptly. This ensures that the map always matches the actual environment and provides a reliable basis for AGV obstacle avoidance path planning.

[0065] S3: Using the Dijkstra algorithm with pruning optimization, an initial path is generated based on the set start and end points on the constructed map. However, given the complexity and real-time requirements of the warehouse environment, these algorithms need to be improved and optimized. By improving the heuristic function, it can better meet the comprehensive considerations of obstacle avoidance and efficiency in the warehouse scenario, guiding the search direction towards a more optimal and safer path.

[0066] S31: Put the starting point into the visited node set, initialize a distance array, set the distance from the starting point to itself as the leader, initialize the distance from the starting point to all other nodes to infinity, and record the predecessor node of each node as empty.

[0067] S32: Traverse all adjacent nodes of the starting point, update the distance values ​​of the corresponding adjacent nodes in the distance array according to the weights of the edges from the starting point to the adjacent nodes, and set the predecessor nodes of these adjacent nodes as the starting point.

[0068] S33: Select the node with the smallest distance from the unvisited nodes to add to the visited set and update its adjacent node distance and predecessor node.

[0069] S331: Select the node with the smallest distance value from the nodes that have not been visited, and add it to the set of visited nodes.

[0070] S332: Recheck and update the distance from all adjacent nodes of the node that has just been added to the visited set to the starting point, and determine whether the distance from the node that has just been added to the visited set to its adjacent node is smaller than the previously recorded distance. If so, update the distance value and update the predecessor node of the corresponding adjacent node to the node that has just been added to the visited set.

[0071] S34: Repeat the operation of selecting the node with the smallest distance from the unvisited nodes to add to the visited set and updating the distance between its adjacent nodes and the predecessor node until the end point is added to the visited node set.

[0072] S35: When the end point enters the visited node set, start from the end point and keep backtracking along the predecessor node to obtain each node from the starting point to the end point in sequence, combine these nodes in order to obtain the initial path.

[0073] S4: Set the repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path.

[0074] S41: Refer to the idea of ​​artificial potential field method and define corresponding repulsion functions for different types of obstacles.

[0075] The repulsion function formula is expressed as:

[0076] In the formula, k rep is the repulsion coefficient, which is used to adjust the magnitude of the repulsion, d is the distance between the AGV and the obstacle i, and ρ is the radius of the obstacle’s influence range. is the vector pointing from the obstacle to the AGV, , the obstacle coordinates are (x obs ,y obs ), AGV coordinates are (x agv ,y agv ).

[0077] S42: Calculate the direction of the resultant repulsive force generated by each obstacle, using the direction of the line connecting the target and the AGV as the direction of gravity. Superimpose the repulsive force vectors calculated for each obstacle based on its repulsive function to obtain a total repulsive force vector, which is then synthesized with the gravity vector to obtain the final resultant force direction on the AGV. This guides the AGV's path planning, allowing it to automatically avoid areas with obstacles. For example, if the target is directly in front of the AGV and the gravity direction is forward, but there is an obstacle on the right that generates a repulsive force to the right, after vector synthesis, the AGV's path will be adjusted in a direction that both approaches the target and avoids the obstacle on the right.

[0078] The calculation formula of the repulsive force vector of each obstacle i is expressed as:

[0079] in, =d is the distance between AGV and obstacle i, and are the components of the repulsive force vector in the x-axis and y-axis directions respectively.

[0080]

[0081]

[0082] Where, is the repulsive force function, the obstacle coordinates are (x obs ,y obs ), AGV coordinates are (x agv ,y agv ).

[0083] The formula for the total repulsive force vector is:

[0084] Where n is the total number of obstacles.

[0085] The total repulsive force vector is combined with the gravitational vector to obtain the final resultant force vector Ftotal on the AGV:

[0086] Where, is the gravitational vector, .

[0087] S43: As the AGV moves and the position of obstacles changes, the distance, relative direction and other parameters between each obstacle and the AGV are continuously updated, and the repulsion function value and the direction of the resultant force are dynamically recalculated to ensure that the path planning can adapt to environmental changes in real time and always guide the AGV to avoid detected obstacles.

[0088] S5: Set a safety buffer zone so that the first optimized path can cope with the slight deviation during the AGV driving process according to the safety buffer zone and obtain the second optimized path.

[0089] S51: Determine the range of the safety buffer zone based on the warehouse data, and set it around the primary optimization path, so that the safety buffer zone is closely associated with the primary optimization path, and the primary optimization path becomes a reference basis for setting the safety buffer zone.

[0090] S52: When the AGV is traveling along an optimized path, its position changes are monitored in real time. The actual driving position monitored is compared with the optimized path and the set safety buffer zone boundary to determine whether there are slight deviations. The optimized path and safety buffer zone are the key references for determining deviations.

[0091] S53: Once a slight deviation is detected and the deviation is within the safety buffer zone, the path is fine-tuned based on the specific direction, distance and other information of the deviation, combined with the range limit of the safety buffer zone and the original direction of the first optimization path. By continuously making such dynamic adjustments based on the deviation, a second optimization path that can cope with slight deviations is finally obtained. The first optimization path, safety buffer zone and monitored deviation conditions involved in the previous steps are all important input bases for obtaining the second optimization path.

[0092] S6: Monitor the AGV's operating status in real time to determine whether the environment has changed. If so, recalculate the new obstacle avoidance path based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environmental map.

[0093] The AGV's surrounding environment is continuously monitored through sensors, and a high-frequency monitoring frequency is set. Once a change in the environment is detected, such as the appearance of a new obstacle, other AGVs changing their routes, or the originally accessible channel being blocked, the path adjustment process is immediately triggered. When the path adjustment demand is triggered, the optimized path planning algorithm is used to quickly recalculate a feasible obstacle avoidance path with the current AGV position as the new starting point and the target point unchanged, combined with the updated environmental map. This recalculation process must be completed in a very short time to ensure that the AGV can follow the new path in a timely manner and avoid safety accidents such as collisions. The environmental map is updated immediately based on the latest sensor data. If a grid map is used, the grid corresponding to the newly appeared obstacle is marked as impassable, and the impassable mark corresponding to the disappeared obstacle is removed. At the same time, the location information of other AGVs and the accessibility status of the channel are updated to ensure that the environmental map can reflect the current actual warehouse situation in real time.

[0094] The actual position of the current AGV is clearly used as the starting point for new path planning, and the AGV's own positioning system is used to accurately obtain its coordinate position in the environmental map.

[0095] If the AGV's mission objective remains unchanged due to environmental changes, the original target point remains unchanged. If the mission objective changes due to special circumstances, a new target point is determined according to the new mission instructions, and the target point identifier in the environmental map is updated at the same time.

[0096] The rationality of the generated path is verified to check whether the path crosses inaccessible areas or conflicts with other AGV planned paths. If there are any unreasonable issues, the algorithm parameters are adjusted in time or an alternative path optimization strategy is used to correct them until a feasible obstacle avoidance path that meets the requirements is obtained. The recalculated obstacle avoidance path information is passed to the AGV control system. The path information includes detailed instructions such as the coordinates of each path node, steering angle, and expected speed, so that the AGV control system can accurately understand and execute the new path planning. Based on the path information, the control system adjusts the AGV's drive wheel speed, steering mechanism, etc. in real time to ensure that the AGV can quickly follow the new path. At the same time, it continues to monitor the environment and prepares to respond to subsequent possible environmental changes. The above environmental monitoring and path adjustment process is executed cyclically to ensure the continuous, safe, and efficient operation of the AGV in the warehouse environment.

[0097] In summary, this embodiment provides an AGV obstacle avoidance path planning method for intelligent warehousing. By using the Dijkstra algorithm with pruning optimization to construct an initial path and setting a repulsion factor and a safety buffer zone to optimize the initial path, the following beneficial effects are achieved: A topological map method focuses on capturing the connections between locations in the environment and between them. By rationally identifying nodes, constructing edges, and assigning attributes, combined with a dynamic update mechanism, the resulting map accurately reflects the real-time state of the warehouse environment, providing realistic and timely foundational data for path planning, facilitating more accurate and optimal path planning. During path planning, an improved Dijkstra algorithm with pruning optimization is employed. This algorithm comprehensively considers obstacle avoidance and efficiency in the warehouse scenario to guide the search direction and quickly determine an initial path. By introducing a repulsive force factor and referencing the artificial potential field method, the path is automatically optimized to avoid obstacles based on a defined repulsive force function and resultant force calculation. A safety buffer is then established, and by monitoring minor deviations and fine-tuning the path, a secondary optimized path is obtained. This series of optimizations results in a safer, more efficient, and adaptable path to the complexities of the warehouse environment. During path planning, an improved Dijkstra algorithm with pruning optimization is employed to comprehensively consider obstacle avoidance and efficiency in the warehouse scenario to guide the search direction and quickly determine an initial path. By introducing repulsive forces and referring to the artificial potential field method, the path is optimized to automatically avoid obstacles based on a defined repulsive force function and resultant force calculation. A safety buffer zone is established, and by monitoring minor deviations and fine-tuning the path, a secondary optimized path is obtained. This series of optimizations makes the planned path safer, more efficient, and more adaptable to complex warehouse environments.

[0098] Based on the same general inventive concept, the present invention also protects an AGV obstacle avoidance path planning system for intelligent warehousing. The following describes an AGV obstacle avoidance path planning system for intelligent warehousing provided by the present invention. The AGV obstacle avoidance path planning system for intelligent warehousing described below and the AGV obstacle avoidance path planning method for intelligent warehousing described above can be referenced to each other.

[0099] Figure 2 This is a module diagram of an AGV obstacle avoidance path planning system for intelligent warehousing provided by an embodiment of the present invention.

[0100] like Figure 2 As shown, an AGV obstacle avoidance path planning system for intelligent warehousing includes: The data acquisition module is used to scan the warehouse data to obtain the warehouse data.

[0101] The data partitioning module is used to divide the warehouse data into different grids using a topological map and construct a map reflecting the topological structure of the warehouse space.

[0102] The path construction module is used to use the Dijkstra algorithm with pruning optimization to obtain the initial path based on the set starting and ending points on the constructed map.

[0103] The path optimization module is used to set the repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path.

[0104] The secondary optimization module is used to set a safety buffer zone so that the primary optimization path can cope with slight deviations during the AGV's driving process according to the safety buffer zone to obtain a secondary optimization path.

[0105] The real-time update module is used to monitor the AGV's operating status in real time to determine whether the environment has changed. If so, the new obstacle avoidance path is recalculated based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environmental map.

[0106] Through the description of the above embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the above technical solution, or the portion that contributes to the prior art, is embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods of each embodiment or certain portions of the embodiments.

[0107] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will appreciate that modifications may be made to the technical solutions described in the aforementioned embodiments, or that some of the technical features therein may be replaced with equivalents. However, such modifications or replacements do not deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An AGV obstacle avoidance path planning method for intelligent warehousing, characterized in that: include: S1: Scan the storage environment to obtain storage data; S2: Divide the storage data into different grids using a topological map to construct a storage space topological structure map; S3: Generate an initial path based on the set starting point and end point on the topological structure map of the storage space using the Dijkstra algorithm with pruning optimization; S4: setting a repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path; S5: Set a safety buffer zone so that the first optimization path can cope with the slight deviation of the AGV during driving according to the safety buffer zone, and obtain the second optimization path; S6: Monitor the AGV's operating status in real time to determine whether the environment has changed. If so, recalculate a new obstacle avoidance path based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environment map.

2. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1 is characterized in that: In step S1, the warehouse data includes obstacle location data, passable channel data, shelf location data, AGV starting point and target point data, and AGV real-time location data.

3. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: In step S2, the specific steps of constructing the storage space topology map are: S21: Analyze the intersection nodes of aisles between shelves and the connection positions between different functional areas based on the warehouse data; S22: Determine whether there is a passable passage between the nodes based on the intersection nodes and the connection positions. If yes, construct an edge between the two nodes and assign edge attribute characteristics to each edge. S23: Integrate the nodes and edges to obtain a map framework of a spatial topological structure; S24: Establish a dynamic update mechanism to update the map framework of the initial spatial topology structure in real time.

4. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: In step S3, the specific steps of using the Dijkstra algorithm with pruning optimization to find a path are as follows: S31: Put the starting point into the visited node set, initialize a distance array, set the distance from the starting point to itself as the leader, initialize the distance from the starting point to all other nodes to infinity, and record the predecessor node of each node as empty; S32: Traverse all adjacent nodes of the starting point, update the distance values ​​of the corresponding adjacent nodes in the distance array according to the weights of the edges from the starting point to the adjacent nodes, and set the predecessor nodes of these adjacent nodes as the starting point; S33: Select the node with the smallest distance from the unvisited nodes and add it to the visited set; S34: Repeat the step of adding the node to the visited set until the end point is added to the visited node set; S35: When the end point is added to the visited node set, start from the end point, backtrack from the predecessor node, obtain each node from the starting point to the end point in sequence, and combine these nodes in order to obtain the initial path.

5. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: In step S33, the specific steps of selecting a node to join the set include: S331: Select the node with the smallest current distance value from the nodes that have not been visited and add it to the set of visited nodes; S332: Recheck and update all adjacent nodes of the node with the minimum current distance value, and determine whether the distance from all adjacent nodes to the starting point is smaller than the distance from the node with the minimum current distance value to the starting point. If so, update the distance value and add the predecessor node of the adjacent node to the node of the visited set.

6. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: The specific steps of path planning adjustment in step S4 are: S41: Define corresponding repulsion functions for different types of obstacles; S42: Calculate the repulsive force vector generated by each obstacle according to the repulsive force function, superimpose the repulsive force vectors of all obstacles to obtain a total repulsive force vector, and synthesize the total repulsive force vector with the attraction vector to obtain the final resultant force direction of the AGV; S43: As the AGV moves and the position of the obstacle changes, the distance parameters between each obstacle and the AGV are updated according to the repulsive force vector; S44: recalculating the repulsion function value according to the distance parameter, and recalculating the resultant force direction according to the repulsion function value to obtain an optimized path.

7. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 6, characterized in that: In step S42, the repulsion function formula is expressed as: Where k rep is the repulsion coefficient, d is the distance between AGV and obstacle i, and ρ is the radius of the obstacle’s influence range; is the vector pointing from the obstacle to the AGV, , the obstacle coordinates are (x obs ,y obs ), AGV coordinates are (x agv ,y agv ).

8. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: In step S5, the specific steps for dealing with minor deviations during AGV driving are as follows: S51: Determine the range of the safety buffer zone according to the storage data; S52: Compare the planned position of the primary optimized path with the range of the safety buffer zone to determine whether the AGV's driving has deviated. If so, adjust the primary optimized path according to the deviation and the range of the safety buffer zone to obtain a secondary optimized path.

9. The AGV obstacle avoidance path planning method for intelligent warehousing according to claim 1, characterized in that: It also includes the collaborative operation of multiple AGVs, dynamically adjusting the paths of each AGV through a conflict-based search algorithm, independently planning a preliminary path for each AGV, and determining whether there are conflicts between the paths through detection. If so, coordination and replanning will be carried out.

10. An AGV obstacle avoidance path planning system for intelligent warehousing, which is applied to an AGV obstacle avoidance path planning method for intelligent warehousing as claimed in any one of claims 1 to 9, characterized in that: The path planning system includes: Data acquisition module, used to scan warehouse data to obtain warehouse data; A data partitioning module is used to divide the storage data into different grids using a topological map to construct a map reflecting the topological structure of the storage space; The path construction module is used to obtain the initial path based on the set starting and ending points on the constructed map using the Dijkstra algorithm with pruning optimization; A path optimization module is used to set a repulsion factor so that the initial path automatically avoids the area where the obstacle is located according to the repulsion factor to obtain an optimized path; The secondary optimization module is used to set a safety buffer zone so that the primary optimization path can cope with slight deviations during the AGV's driving process according to the safety buffer zone, and obtain a secondary optimization path; The real-time update module is used to monitor the AGV's operating status in real time to determine whether the environment has changed. If so, the new obstacle avoidance path is recalculated based on the secondary optimization path, with the current AGV's position as the new starting point and the target point unchanged, combined with the updated environment map.

Images