AGV (Automatic Guided Vehicle) path planning method and device capable of avoiding in advance

Abstract

The invention discloses an AGV path planning method and device capable of avoiding in advance, and relates to the technical field of AGVs, and the method comprises the steps: determining an initial driving path of the AGV according to a current task when the AGV receives the task, and controlling the AGV to drive according to the initial driving path; in the driving process of the AGV, obstacle information collected by an obstacle detection device arranged on a moving path of the AGV is obtained in real time; and when the obstacle is detected, adjusting the initial driving path according to the position information of the detected obstacle, the initial driving path of the AGV trolley and the current position of the AGV trolley so as to avoid the obstacle on the initial driving path of the AGV trolley. The problem that in the prior art, the efficiency of task completion is low due to the fact that an AGV cannot avoid obstacles in advance in the driving process is solved.

Description

Technical Field

[0001] This invention relates to the field of AGV (Automated Guided Vehicle) technology, and in particular to a path planning method and device for AGVs that can avoid obstacles in advance. Background Technology

[0002] With the rapid development of intelligent manufacturing, intelligent warehousing, and logistics automation, Automated Guided Vehicles (AGVs), as core equipment for material handling and process integration, directly determine the turnover efficiency and operating costs of the overall production or warehousing system based on their operating efficiency, safety, and environmental adaptability. Path planning technology, as a core functional module of AGVs, plays a crucial role in "planning a collision-free and efficient path for the AGV from the task's starting point to the target point," and is the foundation for ensuring that AGVs can operate autonomously without human intervention.

[0003] Currently, AGVs plan their travel path in advance before performing a task and travel along the planned path. However, if an obstacle suddenly appears on the path during the journey, and there is no alternative path to choose from at the location of the obstacle, the AGV will be forced to stop moving, affecting the efficiency of the AGV in completing the task. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an AGV path planning method and device that can avoid obstacles in advance, in order to solve the problem of low efficiency in task completion caused by the inability of AGVs to avoid obstacles in advance during the driving process.

[0005] One object of the present invention is to provide a path planning method for AGVs that can avoid obstacles in advance, the method comprising: When the AGV receives a task, it determines the initial driving path of the AGV based on the current task, and controls the AGV to drive according to the initial driving path. During the AGV's movement, obstacle information is collected in real time by obstacle detection devices deployed along the AGV's path. When an obstacle is detected, the initial travel path of the AGV is adjusted based on the location information of the detected obstacle, the initial travel path of the AGV, and the current position of the AGV, in order to avoid the obstacle on the initial travel path of the AGV.

[0006] Furthermore, in the aforementioned AGV path planning method that allows for advance obstacle avoidance, the step of adjusting the initial travel path based on the detected obstacle information, the AGV's initial travel path, and the AGV's current position when an obstacle is detected includes: By combining the location information of the obstacle, the initial travel path of the AGV, and the current position of the AGV, it is determined whether the obstacle appears on the remaining travel path of the AGV. When it is determined that the obstacle does not appear on the remaining driving path of the AGV; Then control the AGV to continue traveling along the initial path; When an obstacle is detected in the remaining travel path of the AGV, the remaining travel path of the AGV is adjusted to obtain the target travel path, and the AGV is controlled to continue traveling along the target travel path.

[0007] Furthermore, in the aforementioned AGV path planning method that allows for advance obstacle avoidance, the step of adjusting the remaining travel path of the AGV to obtain the target travel path includes: The location information of the detected obstacles is updated in the movement path map of the AGV, and the area where the obstacle is located is marked as a temporarily impassable area; Using the current position of the AGV as the starting point for path replanning and the final target point of the initial travel path as the endpoint, a target travel path without obstacles is replanned from the starting point to the endpoint on the updated travel path map using a preset path planning algorithm.

[0008] Furthermore, in the aforementioned AGV path planning method that allows for advance obstacle avoidance, the step of replanning a target travel path from the starting point to the end point without passing through obstacles on the updated travel path map using a preset path planning algorithm includes: Set weighting coefficients to guide the search direction; Define the method for calculating path cost, and the cost function shall include at least path length, travel time, energy consumption, and path smoothness; Based on the updated movement path map, a preset path planning algorithm is invoked with the goal of minimizing path cost. Starting from the origin, the algorithm searches the movement path map with set weight coefficients until a complete path connecting to the destination is found. During the search, avoid all areas marked as impassable, including both newly appearing obstacles and existing obstacles; The target driving path is obtained by smoothing the complete path found through the search.

[0009] Furthermore, in the aforementioned AGV path planning method for advance obstacle avoidance, the step of setting the weight coefficients used to guide the search direction includes: The weight coefficients of each factor in the path cost function are dynamically adjusted based on the current task type, current speed, remaining battery power, and the position and movement trend of obstacles.

[0010] Furthermore, the aforementioned AGV path planning method for advance obstacle avoidance, wherein the step of invoking a preset path planning algorithm with the objective of minimizing path cost, starting from the starting point, searching the movement path map with set weight coefficients until a complete path connecting to the destination is found includes: If a complete path from the starting point to the ending point cannot be found within the preset time threshold or number of iterations, the path planning is deemed to have failed. When path planning fails, the AGV explores a localized path near its current position to find a reachable intermediate target point that is far from the obstacle. Once the intermediate target point is reached, path planning is triggered again; or... Control the AGV to wait briefly at its current position or in a safe area, continuously monitor obstacle status, and retry path planning after a preset time; or After sending assistance requests to other AGVs to obtain more comprehensive environmental information or coordination and scheduling instructions, path planning is then performed.

[0011] Furthermore, in the aforementioned AGV path planning method that allows for advance obstacle avoidance, the step of dynamically adjusting the weight coefficients of each factor in the path cost function based on the AGV's current task type, current speed, remaining battery power, and the position and movement trend of obstacles includes: A set of baseline weighting coefficients is preset, corresponding to path length, travel time, energy consumption, and path smoothness, respectively; Detect the current task type of the AGV (Automated Guided Vehicle). If the task type is an emergency task, then the weighting factor of travel time will be increased on the basis of the base weight. If the task type is an energy-saving task, then the weighting coefficient of energy consumption will be increased on the basis of the base weight. Get the current remaining battery power of the AGV vehicle; If the remaining battery power is lower than the preset low power threshold, the weighting factor of energy consumption will be increased and the weighting factor of driving time will be decreased. Get the current speed of the AGV; If the current driving speed is higher than the preset high speed threshold, the weighting coefficient of path smoothness will be increased. Analyze the location and movement trends of obstacles; If the obstacle is close to the AGV or tends to move closer to the AGV, the weighting coefficient of path smoothness is increased. The weight coefficients adjusted by the above steps are normalized to ensure that the sum of each weight coefficient is 1 or a fixed value. The normalized final weight coefficients are then applied to the path cost function as the optimization target for the next path planning.

[0012] Another object of the present invention is to provide an AGV (Automated Guided Vehicle) path planning device capable of early obstacle avoidance, the device comprising: The control module is used to determine the initial driving path of the AGV based on the current task when the AGV receives a task, and to control the AGV to drive according to the initial driving path. The acquisition module is used to acquire obstacle information collected by the obstacle detection device deployed on the AGV's moving path in real time during the AGV's movement. The adjustment module is used to adjust the initial travel path of the AGV when an obstacle is detected, based on the location information of the detected obstacle, the initial travel path of the AGV, and the current position of the AGV, so as to avoid the obstacle on the initial travel path of the AGV.

[0013] Another object of the present invention is to provide a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0014] Another object of the present invention is to provide an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method described above.

[0015] This invention addresses the problem of low task completion efficiency caused by the inability of AGVs to avoid obstacles during operation. When an AGV receives a task, it determines its initial travel path based on the task and controls the AGV's movement accordingly. During travel, it continuously acquires obstacle information from obstacle detection devices deployed along the AGV's path. When an obstacle is detected, the initial travel path is adjusted based on the obstacle's location, the AGV's initial path, and its current position to avoid the obstacle. In other words, it automatically adjusts the planned path when encountering obstacles, thus avoiding them in advance. This solves the problem of low task completion efficiency caused by the inability of existing AGVs to avoid obstacles in advance. Attached Figure Description

[0016] Figure 1 A flowchart of an AGV path planning method for advance obstacle avoidance provided in an embodiment of the present invention; Figure 2This is a structural block diagram of the AGV path planning device that can avoid obstacles in advance, as shown in the third embodiment of the present invention.

[0017] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0018] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0019] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0021] Example 1 Please see Figure 1 The figure shows the AGV path planning method with advance avoidance in the first embodiment of the present invention, the method including steps S10 to S12.

[0022] Step S10: When the AGV receives a task, the initial driving path of the AGV is determined according to the current task, and the AGV is controlled to drive according to the initial driving path.

[0023] When the AGV receives a task from the central control system, host computer, or scheduling platform via a communication module (such as a wireless LAN or industrial Ethernet), it first parses the key information contained in the task. This information includes at least the starting point (the AGV's current initial position or designated standby position), the destination (the target location for material handling, such as a shelf, production line station, or unloading point), and may also include the task type (such as an emergency replenishment task or a regular material transfer task), the load status (such as no load, full load, and load weight), and the allowable travel speed range. Subsequently, the AGV will call its stored local environment map (which has pre-marked fixed obstacles in the warehouse / workshop, such as pillars, fixed shelves, equipment bases, as well as preset feasible passages, node coordinates and other static environment information). Combined with the starting point and ending point information obtained from task parsing, it will start the preset static path planning algorithm. The algorithm will search and calculate the optimal path from the task starting point to the ending point in the local environment map that meets the basic constraints (such as no fixed obstacles blocking the way, relatively short path length, and compliance with the basic driving capabilities of the AGV). This path is the initial driving path of the AGV to perform this task. The initial driving path is usually stored in the AGV's control unit in the form of a series of continuous coordinate point sequences or polyline trajectories. After the initial driving path is determined, the AGV's control unit will convert the path into specific driving control commands. The drive system (such as motors and servos) will control the AGV's wheel steering and speed. At the same time, the AGV will use its onboard positioning sensors (such as lidar, vision sensors, and encoders) to collect its current position coordinates and attitude information in real time and compare them with the preset coordinate points in the initial driving path. If there is a positional deviation (such as driving deviation caused by ground friction or load changes), the control unit will adjust the drive commands in time to correct the driving direction and speed, ensuring that the AGV can travel stably and accurately to the task destination along the initial driving path. Until new commands are received or environmental changes are detected (such as the appearance of obstacles), the initial driving path will be used as the core driving basis.

[0024] Step S11: During the AGV's movement, obtain obstacle information collected in real time by obstacle detection devices deployed on the AGV's movement path.

[0025] Specifically, during the entire dynamic driving process of the AGV moving from the starting point of the task along the initial driving path to the destination, its control system continuously acquires environmental data collected by multiple obstacle detection devices through a preset communication interface (such as CAN bus, industrial Ethernet) at a high sampling frequency of milliseconds or seconds. These detection devices are not randomly deployed, but are precisely installed on the driving path in the working environment of the AGV according to the coverage requirements of the AGV's moving path. LiDAR and vision cameras can be installed at key points of the moving path, such as forks, to cover obstacles that may appear when turning or on the side path, or infrared sensors can be added. The detection range of all devices has been calibrated to ensure complete coverage of the AGV's driving trajectory. More specifically, in order to achieve the core objective of enabling AGVs to detect obstacles in advance and adjust their paths in a timely manner during movement, the physical layout of the obstacle detection device is designed around blind-spot-free coverage, long-distance early warning, and adaptability to different path scenarios. The communication method must meet the requirements of low latency, high reliability, and large-capacity data transmission. The two work together to ensure that obstacle information can be transmitted to the AGV control system in real time and accurately.

[0026] In terms of physical layout, LiDAR can be selected as the core long-range detection device. 2D / 3D LiDAR, with its resistance to strong light and dust interference, can accurately output the three-dimensional coordinates, size, and movement trend of obstacles. A vision camera serves as an auxiliary recognition device, using industrial-grade high-definition color / infrared cameras to identify obstacle types such as personnel, other AGVs, and scattered goods, supplementing LiDAR's limitation in recognizing "non-rigid obstacles." Millimeter-wave radar serves as a backup device for harsh environments, using 24GHz / 77GHz millimeter-wave radar with a detection range of 5-20m, capable of penetrating fog, dust, rain, and other harsh environments to avoid detection failure in extreme scenarios. Ultrasonic sensors serve as a short-range blind spot compensation device, with a detection range of 0.3-5m, installed in close-range areas on both sides of the path and around loading and unloading points to avoid missed detection of close-range obstacles due to obstructed views from fixed devices, such as gaps between shelves.

[0027] The topology based on the AGV movement path includes straight sections, turns, intersections, and loading / unloading points. Fixed devices are arranged according to the principle of "layered coverage and key point densification" to ensure the detection range covers the entire path without blind spots: For straight sections, LiDAR and vision cameras are installed on the pillars or shelf pillars on both sides of the path, spaced 8-12m apart to match the LiDAR detection range, and 1.2-1.5m high to avoid obstruction from ground debris and cover the AGV's travel height range. The detection range of adjacent devices overlaps by ≥2m to avoid local blind spots caused by device malfunctions. At turns, an additional 3D LiDAR is installed inside the turning radius, such as at the inner corner of a 90° turn, with a detection angle covering the extended section of the path after the turn by ≥15m. Vision cameras are installed on the outside of curves to capture sudden obstacles such as people crossing the path after the turn in real time, ensuring that AGVs can detect obstacles after the turn before entering the turn. At intersections, one 360° lidar is installed at each of the four corners or key locations where paths intersect, with a detection range covering the extension of all intersecting paths ≥20m. Millimeter-wave radar is also deployed as a backup to avoid detection delays caused by signal blockage when multiple AGVs intersect. At loading and unloading points, ultrasonic sensors and vision cameras are installed along the path within a 3-5m radius around the loading and unloading platform to detect goods that may be scattered, temporary personnel, or forklifts during the loading and unloading process, ensuring that AGVs can detect nearby obstacles in advance when approaching the loading and unloading point.

[0028] In terms of communication methods, an "edge computing node + industrial-grade communication network" architecture can be adopted. Obstacle information collected by fixed detection devices is transmitted to the AGV control system in a hierarchical manner, reducing the data transmission link length and thus reducing latency. Edge computing nodes are set up in the installation area of ​​fixed detection devices, such as one edge computing gateway every 20-30m path. The gateway is directly connected to fixed devices within a 50m radius, including LiDAR, cameras, millimeter-wave radar, and ultrasonic sensors, via wired connections. It receives the raw data collected by the devices in real time and performs preliminary processing, such as filtering valid obstacle information and eliminating environmental interference data, to avoid a large amount of invalid data occupying the backbone network bandwidth. In the backbone communication network, the edge computing gateway communicates with the AGV control system and the AGV vehicle controller via a dual-link communication of "industrial Ethernet + wireless backup" to ensure uninterrupted data transmission. The specific communication technology selection adapts to different transmission requirements: Communication between fixed devices and the edge gateway uses Industrial Ethernet (Profinet or EtherNet / IP protocol) via shielded twisted-pair cable or fiber optic cable, with a transmission rate of 100Mbps-1Gbps and a latency of ≤10ms. This is suitable for real-time transmission of large-capacity data such as point cloud data and image data from LiDAR and vision cameras, while also supporting device status monitoring such as LiDAR fault alarms. Communication between the edge gateway and the AGV control system primarily uses Industrial Ethernet (ring topology) to ensure that a single node failure does not affect the overall network, with a transmission rate of 1Gbps and a latency of ≤20ms. This is used to transmit global obstacle information such as the distribution of obstacles at intersections, with Wi-Fi as a backup. 6. Wireless communication, specifically the 5GHz band, enhances anti-interference capabilities, with a transmission rate ≥500Mbps and latency ≤30ms, adapting to mobile transmission needs in flexible path scenarios. Obstacle information sharing among multiple AGVs is achieved through V2X (vehicle-to-vehicle) communication technology, based on dedicated short-range communication or the 5G-V2X protocol. This enables real-time sharing of obstacle information between AGVs. For example, if an AGV detects an obstacle via a fixed device, the information can be synchronized to other AGVs within a 100m radius, avoiding path conflicts caused by information asymmetry among multiple AGVs. Through the above-mentioned purely fixed physical deployment and communication method design, the obstacle detection device can achieve obstacle detection "on the entire path, over long distances, and in multiple scenarios," allowing sufficient time for advance path adjustments and truly realizing "early warning and proactive obstacle avoidance."

[0029] The obstacle information acquired includes not only the basic positional parameters of the obstacle (such as its distance relative to the AGV, horizontal angle, and coordinates in the environmental map), but also its physical attributes (such as size and outline shape to determine whether it will block the path) and motion status (such as whether it is stationary, its speed and direction of movement to predict whether it will enter the AGV's driving path). This real-time collected information is transmitted to the AGV's path planning module in an instant, providing core data support for subsequent judgments on whether obstacles affect the remaining driving path and whether the path needs to be adjusted, thus avoiding collision risks or driving interruptions due to lag in environmental perception.

[0030] Step S12: When an obstacle is detected, the initial travel path is adjusted according to the location information of the detected obstacle, the initial travel path of the AGV, and the current position of the AGV, so as to avoid the obstacle on the initial travel path of the AGV.

[0031] When the obstacle detection device detects a signal that there is an obstacle in the driving environment, its control system will immediately initiate a path adjustment process to avoid the obstacle on the initial driving path of the car.

[0032] Specifically, the step of adjusting the initial travel path based on the detected obstacle information, the initial travel path of the AGV, and the current position of the AGV when an obstacle is detected includes: By combining the location information of the obstacle, the initial travel path of the AGV, and the current position of the AGV, it is determined whether the obstacle appears on the remaining travel path of the AGV. When it is determined that the obstacle does not appear on the remaining driving path of the AGV; Then control the AGV to continue traveling along the initial path; When an obstacle is detected in the remaining travel path of the AGV, the remaining travel path of the AGV is adjusted to obtain the target travel path, and the AGV is controlled to continue traveling along the target travel path.

[0033] The detection device collects and analyzes obstacle location information in real time. This information includes not only the precise coordinates of the obstacle in the global environment map (such as X / Y axis coordinates), but also the obstacle's spatial range (such as length, width, and height, to determine whether it will occupy the passage space of the initial driving path) and its positional relationship relative to the AGV (such as the straight-line distance from the AGV's current position, and whether it is located directly in front of or to the side of the AGV's driving direction). The initial driving path data generated by the AGV when the task starts is stored in the form of a continuous coordinate point sequence, which clearly defines the preset driving trajectory from the task start point to the end point. At this time, the system will focus on extracting the remaining initial driving path segment from the "AGV's current position to the task end point" as a benchmark for judging whether the obstacle has an impact. The AGV obtains its own current position information in real time through positioning sensors (such as encoders and laser SLAM). This information ensures that the starting point of the path adjustment is completely consistent with the actual position of the AGV, avoiding the disconnect between the adjusted path and the actual driving state of the AGV due to positioning deviation. By comparing spatial coordinates, it is determined whether the spatial range of the obstacle overlaps with the remaining initial driving path segment (i.e., whether the obstacle will block the AGV from continuing to drive along the initial path). If it is determined that the obstacle does not appear on the remaining driving path of the AGV, the AGV is controlled to continue driving along the initial driving path. If it is determined that there is an overlap, the remaining driving path of the AGV is adjusted to obtain the target driving path, and the AGV is controlled to continue driving along the target driving path.

[0034] In addition, in some optional embodiments of the present invention, a series of precise judgments and calculations are used to determine whether to adjust the path and how to initially plan the adjustment direction. Specifically, by combining the dynamic characteristics of obstacles with the driving status of the AGV itself, more accurate and efficient obstacle avoidance in advance can be achieved.

[0035] Specifically, the first step is to obtain detailed dynamic information about the obstacle, including whether it is static or dynamic. If it is a dynamic obstacle, its real-time speed and direction of movement need to be obtained. At the same time, the size parameters of the obstacle also need to be determined. This information is the basis for subsequent judgments and calculations. For example, the size parameters will affect the definition of the safe distance.

[0036] Next, based on the AGV's current speed, the remaining path distance between it and the obstacle's location, and the obstacle's current position, the estimated time t1 for the AGV to reach the obstacle's location is calculated. This time t1 is the key time point for determining whether the two might meet.

[0037] Different position prediction methods are used for different types of obstacles: if it is a dynamic obstacle, the predicted position of the obstacle at time t1 is calculated based on its acquired speed and direction of movement; if the obstacle is static, its position will remain unchanged at time t1.

[0038] Next, based on the above calculation results, it is determined whether there is a collision risk between the AGV and the obstacle at time t1, that is, whether the distance between their expected positions at time t1 is less than the safety threshold. If the determination result is that there is no collision risk, it means that even if the AGV travels along the original remaining path, it will not collide with the obstacle at time t1. At this time, the AGV is controlled to continue traveling along the original remaining path to avoid unnecessary path adjustments and ensure efficiency. If the determination result is that there is a collision risk, it means that traveling along the original path may result in a collision. At this time, multiple candidate avoidance paths are generated to prepare for selecting the optimal path from these candidate paths, thereby achieving the goal of avoiding obstacles in advance.

[0039] In summary, the AGV path planning method with advance obstacle avoidance in the above embodiments of the present invention determines the initial travel path of the AGV based on the current task when the AGV receives a task, and controls the AGV's movement according to the initial travel path. During the AGV's movement, obstacle information collected in real time by obstacle detection devices deployed along the AGV's path is acquired. When an obstacle is detected, the initial travel path is adjusted based on the detected obstacle's position information, the AGV's initial travel path, and the AGV's current position to avoid the obstacle on the initial travel path. That is, the planned path is automatically adjusted when facing obstacles, thereby avoiding them in advance. This solves the problem of low task completion efficiency caused by the inability of existing AGVs to avoid obstacles in advance during movement.

[0040] Example 2 This embodiment also proposes an AGV path planning method that allows for advance obstacle avoidance. The difference between the AGV path planning method in this embodiment and the one in Embodiment 1 is as follows: The step of adjusting the remaining travel path of the AGV to obtain the target travel path includes: The location information of the detected obstacles is updated in the movement path map of the AGV, and the area where the obstacle is located is marked as a temporarily impassable area; Using the current position of the AGV as the starting point for path replanning and the final target point of the initial travel path as the endpoint, a target travel path without obstacles is replanned from the starting point to the endpoint on the updated travel path map using a preset path planning algorithm.

[0041] When it is necessary to adjust the remaining travel path of the AGV to obtain the target travel path, the AGV control system first updates the obstacle location information collected in real time by obstacle detection devices (such as LiDAR, vision sensors, etc.) to the movement path map stored locally in the AGV. This map not only includes static environmental information such as the original fixed facilities (such as columns, shelves, equipment bases) in the warehouse, workshop and other scenarios, but also supports the real-time writing of dynamic information. At this time, the system will mark the area where the obstacle is located as a temporary impassable area on the map according to the position and size of the obstacle. This marking is different from the original fixed impassable areas (such as walls) on the map. It is temporary (if the obstacle is removed later, the area can be remarked as passable). Its purpose is to clearly inform the path planning algorithm that "this area cannot be passed at present", so as to provide accurate environmental constraints for subsequent planning. Subsequently, the system uses the current physical position of the AGV, obtained in real time through a positioning module (such as laser SLAM or encoder), as the starting point for path replanning (ensuring that the planning starting point is completely consistent with the actual position of the AGV to avoid positioning deviations that would render the planned path invalid). The system uses the preset final target point in the initial travel path (i.e., the endpoint required by the task, such as a designated shelf or workstation) as the endpoint for replanning. On the updated travel path map (which already includes newly marked temporary impassable areas), the system calls preset path planning algorithms (such as A* algorithm, Dijkstra algorithm, RRT algorithm, etc.). The algorithm searches within the traversable areas on the map (i.e., areas not marked as permanently impassable or temporarily impassable), gradually exploring reachable paths from the starting point until a complete path that connects to the endpoint without passing through any impassable areas is found. This path is the adjusted target travel path used to guide the AGV to continue traveling. Through this process, the system ensures that the AGV avoids currently detected obstacles and that it can still reach the task target point, achieving dynamic path adaptation and safe travel.

[0042] Specifically, the step of replanning a target driving path from the starting point to the destination without passing through obstacles on the updated travel path map using a preset path planning algorithm includes: Set weighting coefficients to guide the search direction; Define the method for calculating path cost, and the cost function shall include at least path length, travel time, energy consumption, and path smoothness; Based on the updated movement path map, a preset path planning algorithm is invoked with the goal of minimizing path cost. Starting from the origin, the algorithm searches the movement path map with set weight coefficients until a complete path connecting to the destination is found. During the search, avoid all areas marked as impassable, including both newly appearing obstacles and existing obstacles; The target driving path is obtained by smoothing the complete path found through the search.

[0043] In practical implementation, firstly, a set of weight coefficients is set to guide the search direction of the algorithm. These coefficients are used to adjust the importance of each evaluation index in the path cost function, ensuring that the algorithm can give priority to the more critical factors in the current scenario when searching. Next, the calculation method of path cost is clarified, and a comprehensive cost function is constructed. This function includes at least four core evaluation indicators: path length (physical distance), travel time (time taken calculated based on the average speed or dynamic speed of the AGV), energy consumption (power consumption during travel, which is related to the path gradient and acceleration frequency), and path smoothness (the rate of change of steering angle in the path, reflecting the smoothness of travel). By multiplying each indicator by its corresponding weight coefficient and summing the results, the comprehensive cost of the path is obtained, thereby quantifying the quality of the path. Subsequently, based on the updated movement path map, the preset path planning algorithm is invoked. The algorithm takes "minimizing the overall cost" as its core objective. Starting from the current position of the AGV (the replanning starting point), it searches within the passable area on the map (i.e., the area not marked as impassable). During the search, the algorithm dynamically adjusts the emphasis on each cost factor according to the pre-set weight coefficients until a continuous path that can completely connect to the final target point of the initial path is found. Throughout the search process, the algorithm strictly avoids all areas marked as impassable, including both temporary impassable areas formed by newly detected obstacles and permanent impassable areas formed by existing fixed obstacles (such as pillars and walls) on the map, ensuring that the planned path has no risk of collision in physical space. Finally, the complete path obtained from the search (usually composed of a series of discrete coordinate points or polylines) is smoothed. Sharp inflection points in the path are eliminated through methods such as B-spline curve fitting and Bézier curve optimization, making the path curvature change continuously and smoothly, adapting to the kinematic constraints of the AGV (such as minimum turning radius and maximum steering angular velocity), and finally forming a target driving path that the AGV can execute smoothly, ensuring that all obstacles are avoided while taking into account the efficiency, economy and feasibility of the path.

[0044] In addition, in some optional embodiments of the present invention, the step of calling a preset path planning algorithm with the goal of minimizing path cost, starting from the starting point, searching the movement path map with set weight coefficients until a complete path connecting to the destination is found includes: If a complete path from the starting point to the ending point cannot be found within the preset time threshold or number of iterations, the path planning is deemed to have failed. When path planning fails, the AGV explores a localized path near its current position to find a reachable intermediate target point that is far from the obstacle. Once the intermediate target point is reached, path planning is triggered again; or... Control the AGV to wait briefly at its current position or in a safe area, continuously monitor obstacle status, and retry path planning after a preset time; or After sending assistance requests to other AGVs to obtain more comprehensive environmental information or coordination and scheduling instructions, path planning is then performed.

[0045] Among them, a fault tolerance mechanism is set up for the path planning algorithm. When the algorithm cannot directly find a complete path from the starting point to the destination, multiple backup strategies are used to ensure that the AGV can continue to advance the task. The specific process is as follows: First, the criteria for judging path planning failure are clarified. The system will pre-set two core thresholds. One is a time threshold, which is the duration from the start of the search to the failure to output a valid path exceeds this value. The other is an iteration number threshold, which is the number of times the algorithm attempts to expand path nodes during the search process exceeds this value. As long as either of these conditions is met, the system will judge it as "path planning failure" to avoid the algorithm getting stuck in an infinite loop and consuming system resources.

[0046] When the planning fails, the system will select the appropriate solution from the following three preset strategies: The first strategy is a local path exploration strategy. The system will define a limited local exploration range (such as a circular area with a radius of 5 meters, to avoid inefficiency due to an excessively large exploration range) centered on the current position of the AGV. Within this range, the system will search for a "reachable and safe intermediate target point"—"reachable" means that there are no impassable areas between the point and the current position of the AGV, and the AGV can drive directly to it; "safe" means that the distance between the point and all surrounding obstacles is not less than a preset safe distance (such as 1 meter, to prevent approaching obstacles). After the AGV reaches the intermediate target point by adjusting its driving, the system will trigger the path planning process again, using the intermediate target point as the new starting point and the original end point as the target point to search for a new path. The system will gradually advance towards the end point through a "segmented approach" method, which is suitable for scenarios where the global path is blocked by multiple obstacles but there is passage space in some areas.

[0047] The second strategy is a wait-and-retry approach. The system first determines whether the AGV's current position is a safe area (such as a preset stop point with no obstacles and a wide field of vision, or no impassable areas within 1 meter of the current position). If not, the system controls the AGV to slowly move to the nearest safe area and then enters a short waiting phase (the waiting time is preset to 3 seconds and can be adjusted according to the probability of obstacle movement). During the waiting period, the system continuously monitors the status of obstacles in the original blocking area through an obstacle detection device (such as determining whether the obstacle has been moved or has deviated from its original position). After the waiting time ends, the system calls the path planning algorithm again to try to search for a path. This strategy is suitable for scenarios where obstacles are temporary (such as temporary staff stops or temporary material stacking) and avoids momentary obstacles by waiting briefly.

[0048] The third strategy is cross-AGV assistance. The system sends assistance requests to other AGVs in the same area via industrial Ethernet or wireless communication modules (such as Wi-Fi 6). The request includes the AGV's current position, target point, and information on detected obstacles. After receiving the request, other AGVs will provide feedback on the global environment information they have (such as clear passages discovered by other AGVs and unmarked passable areas). Some AGVs with scheduling authority will also send coordination instructions (such as "a certain passage is passable, I will temporarily give way to make room"). After obtaining this external information, the AGV updates its local movement path map and restarts the path planning algorithm to search, ensuring continuous operation.

[0049] Furthermore, the step of setting the weight coefficients used to guide the search direction includes: The weight coefficients of each factor in the path cost function are dynamically adjusted based on the current task type, current speed, remaining battery power, and the position and movement trend of obstacles.

[0050] Specifically, the system continuously collects and analyzes the AGV's current task type (e.g., emergency replenishment tasks require priority to ensure timeliness, routine transfer tasks can balance efficiency and energy consumption, and energy-saving tasks require key energy consumption control), current driving speed (e.g., at high speeds, sudden turns and braking should be avoided to ensure stability, while at low speeds, the smoothness requirement can be appropriately reduced), remaining battery power (e.g., in low-battery conditions, low-energy-consumption paths should be prioritized to extend range, while in fully charged conditions, efficiency can be more emphasized), and the real-time status of obstacles (including the distance between obstacles and the AGV—close-range obstacles require more careful avoidance paths, while distant obstacles have less impact; the movement trend of obstacles—if obstacles tend to move closer to the AGV, a smoother and safer avoidance trajectory needs to be planned, while stationary obstacles can be handled according to the conventional obstacle avoidance logic). Based on this real-time information, the system dynamically adjusts the weight coefficients of each cost factor in the comprehensive cost function.

[0051] For example, a set of baseline weight coefficients will be pre-set. These coefficients correspond one-to-one with the four core evaluation factors in the path cost function: path length, travel time, energy consumption, and path smoothness. They serve as the basic reference values ​​for all subsequent adjustment operations, ensuring that the weight ratio of each factor is in an initial balanced state before adjustment. Next, the task scheduling module detects the type of task currently being performed by the AGV. If it is determined to be an emergency task (such as an emergency replenishment task caused by material shortage on the production line), the weight coefficient corresponding to "travel time" is increased on the basis of the baseline weight, thereby guiding the path planning algorithm to prioritize searching for the shorter path and ensuring the timeliness of the task. If it is determined to be an energy-saving task (such as low-priority material transfer during off-peak hours, or scenarios that need to prioritize ensuring the endurance), the weight coefficient corresponding to "energy consumption" is increased on the basis of the baseline weight, prompting the algorithm to prioritize selecting the path with lower energy consumption (such as reducing uphill and rapid acceleration sections). Subsequently, the remaining battery power of the AGV is obtained through the battery management module. If the battery power is lower than the preset low battery threshold (e.g., 20%, which can be preset according to the AGV's endurance), the weight coefficient of "energy consumption" is further increased, while the weight coefficient of "driving time" is reduced to avoid the AGV losing power midway due to choosing a high-energy-consuming path in pursuit of speed. Next, the current travel speed of the AGV is obtained through the speed sensor. If the speed is higher than the preset high-speed threshold (such as 60% of the AGV's maximum design speed, which can be adjusted according to the safety requirements of the scenario), the weight coefficient of "path smoothness" is increased to reduce sharp turns and zigzags in the path, preventing the AGV from tipping over or the cargo from overturning due to excessively sharp turns during high-speed travel. At the same time, the system combines the data collected by the obstacle detection device to analyze the real-time position of the obstacle (such as whether the straight distance to the AGV is less than the safety threshold, such as 5 meters) and the movement trend (predicting whether the obstacle is approaching the AGV's travel path through continuous frame data). If either the "close distance" or "approaching trend" condition is met, the weight coefficient of "path smoothness" is increased to ensure that the AGV can complete the avoidance with a smooth trajectory and avoid the safety risks caused by emergency turns. After completing all the above targeted adjustments, the four sets of weight coefficients are normalized to avoid the path planning result being overly dominated by the excessive weight of a certain factor, thus ensuring the uniformity of the evaluation dimensions of each factor. Finally, the normalized final weight coefficients are substituted into the path cost function as the core optimization objective of the next path planning algorithm, so that the planned path can not only avoid obstacles, but also accurately match the current task requirements, equipment status and environmental safety requirements.

[0052] In summary, the AGV path planning method with advance obstacle avoidance in the above embodiments of the present invention determines the initial travel path of the AGV based on the current task when the AGV receives a task, and controls the AGV's movement according to the initial travel path. During the AGV's movement, obstacle information collected in real time by obstacle detection devices deployed along the AGV's path is acquired. When an obstacle is detected, the initial travel path is adjusted based on the detected obstacle's position information, the AGV's initial travel path, and the AGV's current position to avoid the obstacle on the initial travel path. That is, the planned path is automatically adjusted when facing obstacles, thereby avoiding them in advance. This solves the problem of low task completion efficiency caused by the inability of existing AGVs to avoid obstacles in advance during movement.

[0053] Example 3 Please see Figure 2 The image shows a path planning device for an AGV (Automated Guided Vehicle) that can avoid obstacles in advance, as proposed in the third embodiment of the present invention. The device includes: The control module 100 is used to determine the initial driving path of the AGV based on the current task when the AGV receives a task, and to control the AGV to drive according to the initial driving path. The acquisition module 200 is used to acquire obstacle information collected by the obstacle detection device deployed on the moving path of the AGV in real time during the AGV's movement. The adjustment module 300 is used to adjust the initial travel path of the AGV vehicle according to the location information of the detected obstacle, the initial travel path of the AGV vehicle, and the current position of the AGV vehicle when an obstacle is detected, so as to avoid the obstacle on the initial travel path of the AGV vehicle.

[0054] The functions or operation steps implemented by the above modules are largely the same as those in the above method embodiments, and will not be repeated here.

[0055] Example 4 In another aspect, the present invention provides a readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the steps of the method described in any one of the above embodiments one to two.

[0056] Example 5 In another aspect, the present invention provides an electronic device, the electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any one of the methods described in embodiments one to two above.

[0057] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0058] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable instructions for implementing logical functions, and can be embodied in any computer-readable storage medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable storage medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0059] More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable storage media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0060] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0061] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0062] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A path planning method for AGVs capable of early obstacle avoidance, characterized in that, The method includes: When the AGV receives a task, it determines the initial driving path of the AGV based on the current task, and controls the AGV to drive according to the initial driving path. During the AGV's movement, obstacle information is collected in real time by obstacle detection devices deployed along the AGV's path. When an obstacle is detected, the initial travel path of the AGV is adjusted based on the location information of the detected obstacle, the initial travel path of the AGV, and the current position of the AGV, in order to avoid the obstacle on the initial travel path of the AGV.

2. The AGV path planning method with advance obstacle avoidance as described in claim 1, characterized in that, The step of adjusting the initial travel path based on the detected obstacle information, the initial travel path of the AGV, and the current position of the AGV when an obstacle is detected includes: By combining the location information of the obstacle, the initial travel path of the AGV, and the current position of the AGV, it is determined whether the obstacle appears on the remaining travel path of the AGV. When it is determined that the obstacle does not appear on the remaining driving path of the AGV; Then control the AGV to continue traveling along the initial path; When an obstacle is detected in the remaining travel path of the AGV, the remaining travel path of the AGV is adjusted to obtain the target travel path, and the AGV is controlled to continue traveling along the target travel path.

3. The AGV path planning method with advance obstacle avoidance as described in claim 2, characterized in that, The step of adjusting the remaining travel path of the AGV to obtain the target travel path includes: The location information of the detected obstacles is updated in the movement path map of the AGV, and the area where the obstacle is located is marked as a temporarily impassable area; Using the current position of the AGV as the starting point for path replanning and the final target point of the initial travel path as the endpoint, a target travel path without obstacles is replanned from the starting point to the endpoint on the updated travel path map using a preset path planning algorithm.

4. The AGV path planning method with advance obstacle avoidance as described in claim 3, characterized in that, The step of replanning a target driving path from the starting point to the destination without passing through obstacles on the updated movement path map using a preset path planning algorithm includes: Set weighting coefficients to guide the search direction; Define the method for calculating path cost, and the cost function shall include at least path length, travel time, energy consumption, and path smoothness; Based on the updated movement path map, a preset path planning algorithm is invoked with the goal of minimizing path cost. Starting from the origin, the algorithm searches the movement path map with set weight coefficients until a complete path connecting to the destination is found. During the search, avoid all areas marked as impassable, including both newly appearing obstacles and existing obstacles; The target driving path is obtained by smoothing the complete path found through the search.

5. The AGV path planning method with advance obstacle avoidance as described in claim 4, characterized in that, The step of setting the weight coefficients used to guide the search direction includes: The weight coefficients of each factor in the path cost function are dynamically adjusted based on the current task type, current speed, remaining battery power, and the position and movement trend of obstacles.

6. The AGV path planning method with advance obstacle avoidance as described in claim 4, characterized in that, The step of invoking a preset path planning algorithm with the goal of minimizing path cost, starting from the starting point, searching the movement path map with set weight coefficients until a complete path connecting to the destination is found includes: If a complete path from the starting point to the ending point cannot be found within the preset time threshold or number of iterations, the path planning is deemed to have failed. When path planning fails, the AGV explores a localized path near its current position to find a reachable intermediate target point that is far from the obstacle. Once the intermediate target point is reached, path planning is triggered again; or... Control the AGV to wait briefly at its current position or in a safe area, continuously monitor obstacle status, and retry path planning after a preset time; or After sending assistance requests to other AGVs to obtain more comprehensive environmental information or coordination and scheduling instructions, path planning is then performed.

7. The AGV path planning method with advance obstacle avoidance as described in claim 5, characterized in that, The step of dynamically adjusting the weight coefficients of each factor in the path cost function based on the AGV's current task type, current speed, remaining battery power, and the position and movement trend of obstacles includes: A set of baseline weighting coefficients is preset, corresponding to path length, travel time, energy consumption, and path smoothness, respectively; Detect the current task type of the AGV (Automated Guided Vehicle). If the task type is an emergency task, then the weighting factor of travel time will be increased on the basis of the base weight. If the task type is an energy-saving task, then the weighting coefficient of energy consumption will be increased on the basis of the base weight. Get the current remaining battery power of the AGV vehicle; If the remaining battery power is lower than the preset low power threshold, the weighting factor of energy consumption will be increased and the weighting factor of driving time will be decreased. Get the current speed of the AGV; If the current driving speed is higher than the preset high speed threshold, the weighting coefficient of path smoothness will be increased. Analyze the location and movement trends of obstacles; If the obstacle is close to the AGV or tends to move closer to the AGV, the weighting coefficient of path smoothness is increased. The weight coefficients adjusted by the above steps are normalized to ensure that the sum of each weight coefficient is 1 or a fixed value. The normalized final weight coefficients are then applied to the path cost function as the optimization target for the next path planning.

8. A path planning device for AGV vehicles capable of early obstacle avoidance, characterized in that, The device includes: The control module is used to determine the initial driving path of the AGV based on the current task when the AGV receives a task, and to control the AGV to drive according to the initial driving path. The acquisition module is used to acquire obstacle information collected by the obstacle detection device deployed on the AGV's moving path in real time during the AGV's movement. The adjustment module is used to adjust the initial travel path of the AGV when an obstacle is detected, based on the location information of the detected obstacle, the initial travel path of the AGV, and the current position of the AGV, so as to avoid the obstacle on the initial travel path of the AGV.

9. A readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the steps of the method as described in any one of claims 1 to 7.

Images