Straddle double-body type AGV accurate docking method based on dynamic weight correction

Abstract

The invention discloses a straddle double-body type AGV accurate docking method based on dynamic weight correction, and belongs to the field of robot motion driving and control. The implementation method comprises the steps that in the process that a transfer AGV in a dense parking area is in butt joint with a commercial vehicle, a cloud scheduling system determines the position of the commercial vehicle to be transferred through a GPS positioning system, a vehicle taking coarse path is planned according to the position, and the path is sent to the AGV; when the distances between the to-be-transferred commercial vehicle and the commercial vehicles on the two sides of the to-be-transferred commercial vehicle are inconsistent, the AGV can sacrifice certain butt joint precision with the target vehicle, the distance safety of the two sides is guaranteed, and the distance between the AGV and the two sides of the commercial vehicles in the AGV is achieved, so that the vehicle taking function is achieved under the limit condition. The distance difference between the to-be-transferred commercial vehicle and the commercial vehicles on the two sides is detected in real time, the butt joint path between the AGV and the commercial vehicles is dynamically adjusted, it is ensured that safe and accurate butt joint can still be achieved under the condition that the distance difference exists between the AGV and the adjacent commercial vehicles, and the accurate butt joint efficiency is improved by optimizing path planning.

Description

Technical Field

[0001] The invention relates to a precise docking method of a straddle-type twin-body AGV based on dynamic weight correction, and belongs to the technical field of robot motion drive and control. Background Art

[0002] With the rapid development of industrial automation, AGV (Automated Guided Vehicle) has been widely used in warehousing, logistics distribution, production lines and other fields. The main function of AGV is to automatically carry out material handling, transportation, stacking and other operations through preset paths and procedures. Its main advantages are high efficiency, low cost and labor saving. However, in actual applications, the precise docking between AGV and commercial vehicles has always been a key factor affecting operational efficiency and safety.

[0003] When AGVs are docked with commercial vehicles, many complex factors are involved: the distance between the commercial vehicle and AGV, the difference in spacing between commercial vehicles, docking accuracy, etc. Ideally, the docking of AGVs with commercial vehicles should be precise and interference-free, but in actual operation, due to the inconsistent spacing between adjacent commercial vehicles, it is difficult for AGVs to achieve perfect docking with commercial vehicles, which leads to docking errors and even collisions.

[0004] In traditional AGV systems, the docking process of commodity carts can generally be divided into two main stages: the pick-up stage and the return stage. In the pick-up stage, the AGV needs to accurately dock the commodity cart and dock the docking device of the commodity cart with the fixture, boom or other contact equipment of the AGV to ensure the smooth handling of the items. If the docking accuracy is insufficient, the AGV and the commodity cart may not be firmly connected, causing the items to slip or the fixture to loosen during the handling process, which in turn affects the subsequent transportation tasks. The return stage refers to the process in which the AGV returns the commodity cart to the designated location or shelf. At this time, the AGV needs to ensure not only the docking accuracy, but also ensure that the commodity cart is safely parked at the target location. If the commodity cart is not placed accurately, it may affect the warehousing or production process, and even cause a waste of warehouse space, resulting in a decrease in the efficiency of the automation system.

[0005] The docking accuracy requirements in these two stages are extremely strict. Once any slight deviation occurs in these two stages, it will lead to operation failure, and may even damage the equipment, create safety hazards, and cause large economic losses and time delays. Specifically, factors such as the width of the commodity vehicle, the distance between the commodity vehicles on both sides and the AGV, and the distance between the commodity vehicles on both sides and the commodity vehicles to be transferred will affect the docking accuracy. When the distance between the commodity vehicles on both sides and the commodity vehicles to be transferred is inconsistent, it is difficult for the AGV to ensure accurate docking on the same plane, which may lead to operation failure or collision.

[0006] At present, most AGV and commodity vehicle docking technologies are still based on relatively simple distance measurement algorithms and fixed docking modes. In these systems, AGVs will dock according to fixed paths and distance measurement rules, but these systems usually only consider the distance between the commodity vehicle and the AGV, ignoring the possible distance differences between the commodity vehicles on both sides and the commodity vehicles to be transferred. Therefore, when there is an inconsistency in the distance between commodity vehicles, it is difficult for AGVs to achieve accurate docking on the same plane, and collisions or docking failures are prone to occur.

[0007] In addition, most of the existing docking solutions are based on a static way, that is, a series of rules and paths are set at the beginning of the operation, which shows obvious limitations in the environment where the distance between commodity vehicles is large or other uncertain factors. For example, when the distance between commodity vehicles is large or small, the existing system cannot make corresponding flexible adjustments, which affects the execution effect of AGV. On the other hand, because these traditional solutions ignore dynamic adjustment and adaptive capabilities, the AGV system cannot optimize the docking path in real time according to environmental changes. When faced with a complex and changing working environment, the performance of AGV is difficult to be fully utilized, and the reliability and flexibility of the system are poor. Summary of the invention

[0008] In order to solve the problem that AGV is difficult to accurately dock with commercial vehicles due to the difference in vehicle distance during the docking process, the purpose of the present invention is to provide a straddle-type catamaran AGV precise docking method based on dynamic weight correction. When the distance between the commercial vehicle to be transferred and the commercial vehicles on both sides of it is inconsistent during the docking process of the AGV, the AGV will sacrifice a certain docking accuracy with the target vehicle to ensure the safety of the distance on both sides, so as to achieve the system AGV and the distance between the two sides of the internal commercial vehicle to achieve the function of picking up the vehicle under extreme conditions. By real-time detection of the distance difference between the commercial vehicle to be transferred and the commercial vehicles on both sides, the docking path between the AGV and the commercial vehicle is dynamically adjusted to ensure that the AGV can still achieve safe and accurate docking when there is a difference in distance between adjacent commercial vehicles, and improve the efficiency of the entire system by optimizing path planning.

[0009] The present invention discloses a method for precise docking of a straddle-type twin-body AGV based on dynamic weight correction, comprising the following steps:

[0010] Step 1: The cloud dispatching system uses the GPS positioning system to determine the location of the commodity vehicle to be transferred, plans a rough route for picking up the vehicle based on the location, and sends the route to the AGV;

[0011] Step 2: The AGV uses the GPS system to determine its current position in real time; it moves along the rough path of the vehicle pickup in step 1, and stops when the distance between the AGV and the vehicle to be picked up is less than 7m; the AGV uses multiple laser radars to obtain point cloud information of the surrounding environment from multiple angles; and performs range filtering based on the point cloud information to obtain environmental information around the AGV;

[0012] The multiple laser radars are respectively distributed on both sides of the front end, both sides of the rear end, both sides of the inner side of the front end, and both sides of the inner side of the rear end. The acquired point cloud information is respectively referred to as the front end point cloud information, the rear end point cloud information, the front end inner side point cloud information, and the rear end inner side point cloud information. The laser radars on both sides of the front end and both sides of the rear end of the AGV are mainly used to locate the position of the AGV and perceive the surrounding environment information. The laser radars on both sides of the inner side of the front end and both sides of the inner side of the rear end of the AGV are used to determine the position information of the commodity vehicle to be picked up during the docking process.

[0013] Step 3: Divide the front end point cloud information of the AGV into blocks to obtain the point cloud information of the front, rear, left and right sides and inside of the AGV, and perform point cloud clustering on the point cloud information to obtain the distance between the AGV and the left commercial vehicle, the distance between the AGV and the right commercial vehicle, the distance between the AGV and the commercial vehicle to be picked up, the position of each commercial vehicle, and the distance between the commercial vehicle to be picked up and the commercial vehicles on both sides of it;

[0014] Step 4: According to the posture information and distance information obtained in step 3, a dynamic weight adjustment correction method is used to dynamically adjust the AGV target position and posture at the current position of the AGV, and path planning is performed according to the target position and posture. During the dynamic correction process, the minimum distance from the commodity vehicles on both sides must be ensured until the single-line laser radar at the front end of the AGV scans the wheels of the commodity vehicle to be transferred;

[0015] Step 5: Calculate the distance difference between the wheel and the laser radar based on the point cloud information on the inside of the AGV front end, and adjust the heading angle in real time until the AGV reaches the specified position; determine whether the clamping conditions are met based on the point cloud information on the inside of the AGV rear end;

[0016] Step 6: When the clamping conditions are met, the clamping mechanism is controlled to complete the clamping operation, and then the completed clamping work is sent to the cloud scheduling system. The cloud scheduling system publishes the parking position of the commodity vehicle, and the AGV uses the autonomous navigation system to reach the parking position based on the acquired position information;

[0017] Step 7: Use the GPS positioning system to obtain the location coordinates of the parked vehicle, and upload the location coordinates of the parked vehicle to the cloud scheduling system through MQTT communication, update the location of the next vehicle to be transferred, and repeat steps 1 to 6 to move the next vehicle.

[0018] Furthermore, the specific implementation method of step 4 is:

[0019] The dynamic adjustment and correction AGV method described in the precise docking method of a straddled double-body AGV based on dynamic weight correction can achieve that when the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides of it is inconsistent, the AGV sacrifices a certain docking accuracy with the target vehicle to ensure the safety of the distance on both sides, so as to achieve the distance between the system AGV and the two sides of the internal commodity vehicle to achieve the vehicle picking function under extreme conditions; the specific steps are as follows:

[0020] Step 4.1: Real-time measurement and data collection: First, the distance d between the commodity vehicle to be transferred and the left and right sides obtained by the multi-line laser radar 1 and d 2 , calculate the difference Δd between the two sides of the commercial vehicle:

[0021] Δd=|d 1 -d 2 |

[0022] Among them, d 1 >d 2 Indicates that the left side has a larger spacing; d 1 <d 2 Indicates that the spacing on the right is larger; d 1 =d 2 Indicates that the distance between the two sides is equal, and the real-time detection data will serve as the basic input for subsequent dynamic adjustments;

[0023] Step 4.2: Dynamic correction weight calculation: According to the difference Δd between the two sides of the commodity vehicle, the dynamic correction weight W is calculated. The weight W is used to quantify the importance of the distance difference to the AGV path adjustment. The formula is:

[0024]

[0025] Among them: the value range of W is [0,1]; when W is larger, it means that the difference in the distance between the two sides is more obvious, and the AGV needs to make a larger path adjustment; when W is smaller, it means that the distance between the two sides is similar, and the AGV path adjustment range can be reduced;

[0026] Step 4.3: Path planning adjustment; based on the dynamic correction weight W and the real-time measured distance data d 1 With d 2 , the target position P of the AGV after adjustment adjusted , to ensure that the AGV can avoid interference from the commercial vehicles on both sides when docking, and to ensure the minimum safety distance d min ; The target position adjustment formula of AGV is:

[0027] P adjusted =P target +α·(e W -1)·n

[0028] Among them, Padjusted is the target position after AGV adjustment; P target is the target position before adjustment; α is the adjustment coefficient, which controls the adjustment amplitude; W is the dynamic correction weight; n is the unit vector of the adjustment direction;

[0029] Step 4.4: Dynamic feedback and path correction: During the AGV movement, the system needs to monitor the current position P of the AGV in real time. current and the target position P adjusted The deviation ΔP is calculated and the docking accuracy is ensured by the path correction algorithm; the calculation formula of the deviation ΔP is:

[0030] ΔP=P adjusted -P current

[0031] Corrected position P corrected for:

[0032] P corrected =P current +λ·ΔP

[0033] Among them: λ is the correction proportional coefficient, which is used to control the correction amount; P corrected is the corrected AGV position;

[0034] Step 4.5: Use Hybird A* path planning; use the Hybird A* algorithm to plan the AGV path according to the calculated target position of the AGV; when the AGV reaches the adjusted position P corrected During the whole process of picking up the vehicle, the AGV always ensures that the distance between it and the commercial vehicles on both sides meets the minimum distance until the internal laser radar at the front end of the AGV scans the wheel information.

[0035] Through the above method, the present invention can effectively solve the impact of the difference in the distance between the commodity vehicles on both sides on the AGV path planning and docking accuracy, avoiding collision or docking failure caused by uneven distance. At the same time, through real-time feedback and dynamic adjustment, the system efficiency and operational safety are improved.

[0036] Beneficial effects:

[0037] 1. The present invention discloses a precise docking method for a straddle-type twin-body AGV based on dynamic weight correction. When the distance between the to-be-transferred commodity vehicle and the commodity vehicles on both sides of the AGV is inconsistent during the docking process of the commodity vehicle, the AGV will sacrifice a certain degree of docking accuracy with the target vehicle to ensure the safety of the distance between the two sides, so as to achieve the function of picking up the vehicle under extreme conditions by ensuring the distance between the system AGV and the two sides of the internal commodity vehicle. Through this method, the AGV can ensure high-precision docking in the face of complex environments and uneven spacing, avoiding docking failures or collisions caused by spacing differences in traditional solutions, thereby improving the flexibility and reliability of the precise docking system of the straddle-type twin-body AGV.

[0038] 2. The present invention discloses a straddle-type twin-body AGV precise docking method based on dynamic weight correction, which introduces a dynamic weight adjustment mechanism based on real-time detection of the distance difference between the two sides of the commercial vehicle. The AGV can automatically adjust the path according to the real-time changes in the environment and has stronger adaptive capabilities. Through this adaptive optimization, the AGV can handle more complex operating scenarios, improve the intelligence level of the straddle-type twin-body AGV precise docking system, and avoid the traditional AGV docking system that usually relies on fixed path planning and lacks the ability to respond to dynamic changes in the environment in real time.

[0039] 3. The invention discloses a precise docking method for straddling double-body AGV based on dynamic weight correction, which not only optimizes the docking accuracy between AGV and commodity vehicle, but also avoids unnecessary adjustment and retreat caused by path errors by dynamically adjusting the path, thus reducing possible delays in the handling process. In addition, the precise docking between AGV and commodity vehicle reduces the risk of collision and equipment damage, and reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flow chart of a method for precise docking of a straddle-type twin-body AGV based on dynamic weight correction of the present invention;

[0041] Figure 2 It is a schematic diagram of the straddle-type double-body vehicle structure;

[0042] Figure 3 It is a schematic diagram of the core idea of ​​an AGV precise docking and path optimization method based on dynamic weight correction proposed in the present invention. DETAILED DESCRIPTION

[0043] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0044] This embodiment discloses a precise docking method for a straddling double-body AGV based on dynamic weight correction, which realizes that when the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides is inconsistent during the docking process of the AGV in the dense parking area, the AGV will sacrifice a certain docking accuracy with the target vehicle to ensure the safety of the distance on both sides, so as to achieve the distance between the system AGV and the commodity vehicles on both sides to achieve the function of picking up the vehicle in extreme situations. This patent detects the difference in the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides in real time, and dynamically adjusts the docking path between the AGV and the commodity vehicle to achieve the function of picking up the vehicle in extreme situations, while still being able to achieve safe and accurate docking, and improves the efficiency of the entire system by optimizing path planning.

[0045] like Figure 3 As shown in the figure, when the AGV approaches the position of the commodity vehicle to be picked up, it will use a multi-line laser radar and four single-line laser radars carried by the AGV to sense the surrounding environment, and obtain the position of the commodity vehicle to be transferred and the distance d between the commodity vehicles on both sides through the acquired point cloud information. 1 and d 2 Most of the existing docking solutions are based on the static docking method of aligning the central axis of the commodity vehicle to be transferred with the AGV. As a result, when there is a limit distance between the commodity vehicle to be transferred and the commodity vehicle on the side, the AGV and the commodity vehicle cannot be safely docked, and even scratches the surrounding commodity vehicles. First of all, the AGV structure of this patent is a straddle-type double-body vehicle structure. Its structure is as follows Figure 2 As shown, the middle part of its structure is hollow. In this patent, according to the structure of the AGV and considering the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides, a straddling catamaran AGV precise docking method based on dynamic weight correction is proposed, which sacrifices the left and right spacing inside the AGV (that is, the central axis of the AGV will be offset from the central axis of the commodity vehicle to be transferred according to the difference in the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides), ensuring that safe and accurate docking can still be achieved when there is an extreme spacing, and improving the efficiency of the entire system by optimizing path planning. The distance between the two sides of the commodity vehicle to be transferred and the posture of the commodity vehicle to be transferred are dynamically obtained according to the sensors carried by the AGV, and the target position of the AGV is dynamically corrected according to the distance difference and posture information, and the Hybird A* algorithm is used to perform path planning according to the target position. The specific real-time steps are as follows:

[0046] Step 1: After the cloud dispatch system determines the location of the commodity vehicle to be transferred, it will plan the vehicle pickup path based on the position of the straddling catamaran AGV and send the path to the AGV. The AGV will move along the path toward the commodity vehicle to be transferred using autonomous navigation technology;

[0047] Step 2: The AGV uses the GPS system to determine its current position in real time; it moves along the rough path of the vehicle pickup in step 1, and stops when the distance between the AGV and the vehicle to be picked up is less than 7m; the AGV uses multiple laser radars to obtain point cloud information of the surrounding environment from multiple angles; and performs range filtering based on the point cloud information to obtain environmental information around the AGV;

[0048] The multiple laser radars are respectively distributed on both sides of the front end, both sides of the rear end, both sides of the inner side of the front end, and both sides of the inner side of the rear end. The acquired point cloud information is respectively referred to as the front end point cloud information, the rear end point cloud information, the front end inner side point cloud information, and the rear end inner side point cloud information. The laser radars on both sides of the front end and both sides of the rear end of the AGV are mainly used to locate the position of the AGV and perceive the surrounding environment information. The laser radars on both sides of the inner side of the front end and both sides of the inner side of the rear end of the AGV are used to determine the position information of the commodity vehicle to be picked up during the docking process.

[0049] Step 3: Perform range filtering based on the point cloud data to obtain the environmental information around the AGV. At the same time, the 32-line laser radars on both sides of the front end of the AGV scan the surrounding environment and divide its point cloud information into blocks to obtain the point cloud information in front of, behind, on both sides and inside the AGV. The above point cloud information is clustered and the distances to the left and right commercial vehicles and the commercial vehicles to be picked up, the postures of each commercial vehicle, and the distances between the commercial vehicle to be picked up and the commercial vehicles on both sides of the AGV are determined.

[0050] Step 4: Based on the acquired posture information and distance information, a dynamic weight adjustment correction method is used to dynamically adjust the AGV target position and posture at the current position of the AGV, and path planning is performed based on the target position and posture. During the dynamic correction process, the minimum distance from the commercial vehicles on both sides must be ensured until the single-line laser radar at the front end of the AGV scans the wheels of the commercial vehicle to be transferred.

[0051] Step 5: Collect point cloud data based on the two single-line laser radars at the front end of the AGV and the two correction laser radars at the back end. Calculate the distance difference between the wheel and the laser radar by setting the initial filter range, and adjust the heading angle in real time until the AGV reaches the specified position.

[0052] Step 6: Control the clamping mechanism to complete the clamping operation, and then send the completed clamping work to the cloud scheduling system. The cloud scheduling system publishes the parking location of the commodity vehicle. The AGV uses the autonomous navigation system to reach the parking location based on the acquired location information.

[0053] Step 7: Use the GPS positioning system to obtain the location coordinates of the parked commercial vehicle, and upload the location coordinates of the parked commercial vehicle to the cloud scheduling system through MQTT communication, update the location of the next commercial vehicle to be transferred, and the AGV uses autonomous navigation technology to move the next commercial vehicle.

[0054] A dynamic adjustment and correction AGV method described in a straddle-type twin-body AGV precise docking method based on dynamic weight correction is as follows Figure 1 As shown, the specific implementation steps are as follows:

[0055] Step 4.1: Real-time measurement and data collection. During the implementation process, the AGV needs to first collect environmental data through a variety of sensor devices, including the position of the commodity vehicle to be transferred, the current position of the AGV, and the distance between the commodity vehicle to be transferred and the two sides, and calculate the relative distance between the commodity vehicles.

[0056] Δd=|d 1 -d 2 |

[0057] Among them, d 1 Indicates the distance between the commodity vehicle to be transferred and the commodity vehicle on the left, d 2 Indicates the distance between the commodity vehicle to be transferred and the commodity vehicle on the right.

[0058] Step 4.2: Dynamic correction weight calculation. According to the real-time measurement of the distance d between the two sides 1 and d 2 , the weight W is calculated to quantify the importance of spacing difference to AGV path adjustment, and its formula is:

[0059]

[0060] Among them: the value range of W is [0,1], which indicates the relative degree of the spacing difference; Δd ​​indicates the absolute difference in spacing; d 1 +d 2 It is the total distance between the commodity vehicles on both sides and the commodity vehicles to be transferred.

[0061] When W is larger, it means that the difference in the distance between the two sides is more obvious, and the AGV needs to make a larger path adjustment; when W is smaller, it means that the distance between the two sides is similar, and the AGV path adjustment range can be reduced;

[0062] Step 4.3: Dynamically adjust the AGV path based on the weight. Based on the dynamic correction weight W and the real-time measured distance data d 1 With d 2 , the target position P of the AGV after adjustment adjusted , to ensure that the AGV can avoid interference from the commercial vehicles on both sides when docking, and to ensure the minimum safety distance d min ; The target position adjustment formula of AGV is:

[0063] P adjusted =P target +α·(e W -1)·n

[0064] Among them, P adjustedis the adjusted AGV target position, P target is the target position of the AGV before adjustment. α is an adjustment parameter, which indicates the influence of the dynamic weight on the control amplitude, and n is the unit vector of the adjustment direction;

[0065] Rules for determining the adjustment direction n:

[0066] When 1 >d 2 When , the AGV target position is adjusted to the right, and the direction vector is n = (1,0);

[0067] When 1 <d 2 When , the AGV target position is adjusted to the left, and the direction vector is n = (-1, 0);

[0068] When 1 =d 2 When , no adjustment is required, and the direction vector is n = (0,0);

[0069] Step 4.4: Path correction and dynamic feedback. During the process of AGV picking up or retracting the vehicle, the system needs to continuously detect the relative position of the AGV and the commodity vehicle to ensure the accuracy and real-time performance of the path adjustment. Once a deviation is detected in the docking between the AGV and the commodity vehicle, the AGV control system will make immediate corrections to the path to ensure accurate docking. The real-time correction process is as follows: First, the relative position of the AGV and the commodity vehicle is monitored by laser radar, visual sensors and other equipment to obtain real-time data. If there is a deviation between the AGV and the commodity vehicle, the control system will correct the movement direction and speed of the AGV based on the real-time data to offset it towards the target docking position. The path correction formula is as follows:

[0070] ΔP=P adjusted -P current

[0071] Corrected position P corrected for:

[0072] P corrected =P current +λ·ΔP

[0073] Among them: λ is the correction proportional coefficient, which is used to control the correction amount; P corrected is the corrected AGV position.

[0074] Step 4.5: Use Hybird A* path planning. Use the Hybird A* algorithm to plan the AGV path based on the calculated target position of the AGV. When the AGV reaches the adjusted target position P correctedDuring the whole process, the AGV always ensures that the minimum distance with the commodity vehicles on both sides meets the following conditions: 1 ≥d min =0.2m,d 2 ≥d min = 0.2m, until the laser radar inside the front end of the AGV scans the wheel information.

[0075] In the patent, the Hybrid A* algorithm is an algorithm for robot path planning that combines the heuristic search of the A algorithm with the search method in continuous space. Compared with the classic A* algorithm, Hybrid A* can generate a smoother and more executable path by considering the dynamic constraints of the robot and the actual motion model, especially when the vehicle posture such as rotation and steering needs to be considered. The steps of the Hybrid A* algorithm are:

[0076] (1) Initialization: Determine the starting and ending position information of the AGV, define the robot kinematic model, including the maximum turning radius, maximum speed, etc., use the heuristic function to initialize the open list, and set the obstacle information.

[0077] (2) Heuristic function: Use a heuristic function, such as the straight-line distance from the current position to the end point. When considering the posture factor, it is usually the Euclidean distance plus the angle difference to help guide the search direction. Common heuristic functions are:

[0078] h(n)=sqrt((x_goal-x) 2 +(y_goal-y) 2 )+Δθ

[0079] where Δθ is the difference between the current angle and the target angle.

[0080] (3) Path planning: Use an extended step similar to the A* algorithm to transfer from the current state to the next state through a legal control input (action). Each state transition needs to consider not only the position (x, y) but also the orientation angle, and calculate the cost function for each state.

[0081] (4) Calculation and operation model: When performing state expansion, Hybrid A* will simulate possible control actions based on the robot's motion model. Common motion models include differential drive models or Ackermann models. For each state transition, the robot needs to consider not only the current position, but also the influence of factors such as steering and speed.

[0082] (5) Path correction and smoothing: During the search process, Hybrid A* avoids the robot crossing obstacles and tries to generate a smooth path that does not require a large number of sharp turns. Once a path is found, the path is usually post-processed and smoothed to reduce unnecessary turns to improve execution efficiency and path feasibility.

[0083] (6) When the target state is expanded or found, the algorithm terminates and returns the path. If no path is found after all possible states are expanded, it means that the path planning fails.

[0084] According to the disclosure and teaching of the above description, those skilled in the art to which the present invention belongs may also make changes and modifications to the above embodiments. Therefore, the present invention is not limited to the specific embodiments disclosed and described above, and some modifications and changes to the present invention should also fall within the scope of protection of the claims of the present invention. In addition, although some specific terms are used in this specification, these terms are only for the convenience of description and do not constitute any limitation to the present invention.

Claims

1. A precise docking method for straddling twin-body AGVs based on dynamic weight correction, characterized by: The steps include: Step 1: The cloud dispatching system uses the GPS positioning system to determine the location of the commodity vehicle to be transferred, plans a rough route for picking up the vehicle based on the location, and sends the route to the AGV; Step 2: The AGV uses the GPS system to determine its current position in real time; it moves along the rough path of the vehicle pickup in step 1, and stops when the distance between the AGV and the vehicle to be picked up is less than 7m; the AGV uses multiple laser radars to obtain point cloud information of the surrounding environment from multiple angles; and performs range filtering based on the point cloud information to obtain environmental information around the AGV; The multiple laser radars are respectively distributed on both sides of the front end, both sides of the rear end, both sides of the inner side of the front end, and both sides of the inner side of the rear end. The acquired point cloud information is respectively referred to as the front end point cloud information, the rear end point cloud information, the front end inner side point cloud information, and the rear end inner side point cloud information. The laser radars on both sides of the front end and both sides of the rear end of the AGV are mainly used to locate the position of the AGV and perceive the surrounding environment information. The laser radars on both sides of the inner side of the front end and both sides of the inner side of the rear end of the AGV are used to determine the position information of the commodity vehicle to be picked up during the docking process. Step 3: Divide the front end point cloud information of the AGV into blocks to obtain the point cloud information of the front, rear, left and right sides and inside of the AGV, and perform point cloud clustering on the point cloud information to obtain the distance between the AGV and the left commercial vehicle, the distance between the AGV and the right commercial vehicle, the distance between the AGV and the commercial vehicle to be picked up, the position of each commercial vehicle, and the distance between the commercial vehicle to be picked up and the commercial vehicles on both sides of it; Step 4: According to the posture information and distance information obtained in step 3, a dynamic weight adjustment correction method is used to dynamically adjust the AGV target position and posture at the current position of the AGV, and path planning is performed according to the target position and posture. During the dynamic correction process, the minimum distance from the commodity vehicles on both sides must be ensured until the single-line laser radar at the front end of the AGV scans the wheels of the commodity vehicle to be transferred; Step 5: Calculate the distance difference between the wheel and the laser radar based on the point cloud information on the inside of the AGV front end, and adjust the heading angle in real time until the AGV reaches the specified position; determine whether the clamping conditions are met based on the point cloud information on the inside of the AGV rear end; Step 6: When the clamping conditions are met, the clamping mechanism is controlled to complete the clamping operation, and then the completed clamping work is sent to the cloud scheduling system. The cloud scheduling system publishes the parking position of the commodity vehicle, and the AGV uses the autonomous navigation system to reach the parking position based on the acquired position information; Step 7: Use the GPS positioning system to obtain the location coordinates of the parked vehicle, and upload the location coordinates of the parked vehicle to the cloud scheduling system through MQTT communication, update the location of the next vehicle to be transferred, and repeat steps 1 to 6 to move the next vehicle.

2. The method for precise docking of a straddle-type twin-body AGV based on dynamic weight correction as claimed in claim 1, characterized in that: The specific implementation method of step 4 is: The dynamic adjustment and correction AGV method is used to achieve that when the distance between the commodity vehicle to be transferred and the commodity vehicles on both sides is inconsistent, the AGV sacrifices a certain docking accuracy with the target vehicle to ensure the distance safety on both sides, so as to achieve the distance between the system AGV and the commodity vehicles on both sides to achieve the vehicle picking function under extreme conditions; the specific steps are as follows: Step 4.1: Real-time measurement and data collection: First, based on the distances d1 and d2 between the commodity vehicle to be transferred and the left and right sides obtained by the multi-line laser radar, calculate the difference Δd between the two sides of the commodity vehicle: Δd=|d1-d2| Among them, d1>d2 means that the spacing on the left is larger; d1<d2 means that the spacing on the right is larger; d1=d2 means that the spacing on both sides is equal. The real-time detection data will serve as the basic input for subsequent dynamic adjustment; Step 4.2: Calculate the dynamic correction weight; Calculate the dynamic correction weight W according to the difference Δd between the two sides of the commodity vehicle. The weight W is used to quantify the importance of the distance difference to the AGV path adjustment. The formula is: Among them: the value range of W is [0,1]; when W is larger, it means that the difference in the distance between the two sides is more obvious, and the AGV needs to make a larger path adjustment; when W is smaller, it means that the distance between the two sides is similar, and the AGV path adjustment range can be reduced; Step 4.3: Path planning adjustment: Based on the dynamic correction weight W and the real-time measured distance data d1 and d2, the target position P of the AGV is adjusted. adjusted , to ensure that the AGV can avoid interference from the commercial vehicles on both sides when docking, and to ensure the minimum safety distance d min ; The target position adjustment formula of AGV is: P adjusted =P target +α·(and W -1)·n Among them, P adjusted is the target position after AGV adjustment; P target is the target position before adjustment; α is the adjustment coefficient, which controls the adjustment amplitude; W is the dynamic correction weight; n is the unit vector of the adjustment direction; Rules for determining the adjustment direction n: When d1>d2, the AGV target position is adjusted to the right, and the direction vector is n=(1,0); When d1<d2, the AGV target position is adjusted to the left, and the direction vector is n=(-1,0); When d1=d2, no adjustment is required and the direction vector is n=(0,0); Step 4.4: Dynamic feedback and path correction: During the AGV movement, the current position P of the AGV needs to be monitored in real time. current and the target position P adjusted The deviation ΔP is calculated and the docking accuracy is ensured by the path correction algorithm; the calculation formula of the deviation ΔP is: ΔP=P adjusted -P current Corrected position P corrected for: P corrected =P current +λ·ΔP Among them: λ is the correction proportional coefficient, which is used to control the correction amount; P corrected is the corrected AGV position; Step 4.5: Use Hybird A* path planning; use the Hybird A* algorithm to plan the AGV path according to the calculated target position of the AGV; when the AGV reaches the adjusted position P corrected During the whole process of picking up the vehicle, the AGV always ensures that the distance between it and the commercial vehicles on both sides meets the minimum distance until the internal laser radar at the front end of the AGV scans the wheel information.

Images