Path planning method and system for intelligent loading and stacking robot with body based on segmented path optimization

Abstract

The invention discloses a path planning method and system for an intelligent loading and stacking robot with a body based on segmented path optimization, and the method comprises the steps: obtaining the basic information of a carriage and the basic information of cargos, and setting the stacking priorities of different cargos based on the basic information of the carriage and the basic information of the cargos; in the process of controlling the truck loading and stacking robot to move along the final global path, the expected speed and the expected direction of the truck loading and stacking robot are set for a path formed by adjacent path points; adjacent truck loading stacking robots and the relative position and the relative speed of each adjacent truck loading stacking robot and the current truck loading stacking robot are obtained; the speed and the direction, closest to the expected speed and the expected direction, of the ORCA semi-plane constraint are selected as the final speed and the final direction of the corresponding loading stacking robot; and controlling the loading and stacking robot to stack goods based on the initial local path and the final speed and direction of the loading and stacking robot.

Description

Technical Field

[0001] The present invention belongs to the technical field of path planning, and in particular relates to a path planning method and system for an embodied intelligent loading and stacking robot based on segmented path optimization. Background Art

[0002] In the logistics and warehousing sector, loading and palletizing operations are crucial steps before cargo is shipped. With the advancement of industrial automation, loading and palletizing robots are gradually replacing manual labor in this task, significantly improving operational efficiency and stability. However, existing loading and palletizing robots still face numerous challenges in path planning.

[0003] On the one hand, traditional path planning methods often only consider the shortest path from the robot's starting point to the destination, ignoring factors such as the order in which the goods are placed during the actual loading process, the layout of the compartment space, and the robot's own motion characteristics. This may cause the robot to collide with already stacked goods during movement, repeat paths too many times, and reduce stacking efficiency or even damage the goods. On the other hand, when faced with complex and changing loading tasks, such as goods of different specifications and different compartment sizes, existing path planning methods have poor adaptability and it is difficult to quickly generate reasonable path planning solutions, which cannot meet the efficient and flexible operation requirements in actual production. Existing technologies lack a detailed decomposition of stacking tasks based on cargo priority and compartment layout. The handling of constraints between goods may not be systematic enough, making it difficult to effectively avoid errors in the stacking order. The subtask sorting does not fully consider the compartment space and the robot's motion path, resulting in insufficient operation planning and organization. Summary of the Invention

[0004] The object of the present invention is to provide a path planning method and system for an embodied intelligent loading and stacking robot based on segmented path optimization.

[0005] In a first aspect, the present invention provides a path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization, the method comprising: Obtain the basic information of the carriage and the cargo, and set the stacking priority of different cargoes based on the basic information of the carriage and the cargo; set the stacking coordinates of each cargo according to the stacking priority of the cargo; use the path planning algorithm to obtain the initial local path for each loading and palletizing robot to transport the cargo; for each loading and palletizing robot, obtain multiple global paths by adjusting the corresponding cargo it transports and the order in which the cargo is transported; calculate the total time cost of each global path separately, and select the global path with the lowest total time cost as the optimized global path; use the collision avoidance algorithm based on the optimized global path to obtain the movement direction and speed of the loading and palletizing robot, so as to control the loading and palletizing robot to transport the cargo.

[0006] Preferably, after obtaining the optimized global path, the final global path is obtained by a simulated annealing algorithm.

[0007] As a preferred method, the specific process of obtaining the final global path is as follows: construct a cost function; generate a new path by perturbing the current path, and determine whether to accept the new path based on the Monte Carlo criterion. The acceptance probability formula is: ;in, is the cost function difference between the new path and the current path; is the current temperature; when When , the new path is used as the current path; when When the probability Accept new path; temperature According to the descent rate Gradually reduce, that is ,As the temperature drops, the final global path of the robot carrying the goods is finally obtained.

[0008] Preferably, the cost function The method to obtain is as follows: in, is the path length; Obstacle avoidance cost; The compatibility between the path and the stacking method; is the weight coefficient.

[0009] As a preferred method, the method for obtaining the movement direction and speed of the loading and stacking robot is as follows: Each loading and palletizing robot is modeled as a circle, with the center being the position of the loading and palletizing robot and the radius being the preset safety buffer distance; the expected speed and expected direction of the loading and palletizing robot are set; wherein the expected direction coincides with the direction of the final global path; the loading and palletizing robot whose distance from the current loading and palletizing robot is less than a preset distance threshold is selected as the adjacent loading and palletizing robot of the current loading and palletizing robot; the relative position and relative speed of each adjacent loading and palletizing robot and the current loading and palletizing robot are obtained; the collision avoidance algorithm half-plane constraint of each loading and palletizing robot and the adjacent loading and palletizing robot is obtained based on the relative position and relative speed; the speed and direction of the collision avoidance algorithm half-plane constraint closest to the expected speed and expected direction are selected as the final speed and movement direction of the corresponding loading and palletizing robot.

[0010] Preferably, the basic information of the carriage includes carriage size, carriage shape and obstacle location; the basic information of the cargo includes cargo size, shape and weight.

[0011] Preferably, the method for obtaining the basic information of the carriage is: respectively obtaining the point cloud data, depth and texture information of the loaded carriage, fusing the two types of data after unifying the coordinate system through coordinate conversion technology, and then constructing a three-dimensional spatial model of the carriage with the help of a three-dimensional modeling algorithm.

[0012] In the second aspect, the present invention provides an embodied intelligent loading and palletizing robot path planning system based on segmented path optimization, which is used to execute the above-mentioned embodied intelligent loading and palletizing robot path planning method; the embodied intelligent loading and palletizing robot path planning system includes a data acquisition module, a path generation module, a path optimization module and a controller; the data acquisition module includes a laser radar and a depth camera; the laser radar is used to obtain point cloud data of the carriage; the depth camera is used to capture depth and texture information; the path generation module is used to generate an initial global path for the loading and palletizing robot to palletize goods; and the controller is used to control the loading and palletizing robot according to the acquired path.

[0013] In a third aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the memory stores the computer program; and the processor executes the above-mentioned embodied intelligent loading and palletizing robot path planning method.

[0014] In a fourth aspect, the present invention provides a readable storage medium storing a computer program; when the computer program is executed by a processor, it is used to implement the above-mentioned embodied intelligent loading and palletizing robot path planning method.

[0015] The present invention has the following beneficial effects: 1. The present invention decomposes tasks based on the multi-dimensional characteristics of cargo and the compartment space layout, scientifically allocates cargo stacking locations and determines precise coordinates, and rationally sorts subtasks. This ensures that high-priority cargo is stacked first, and optimizes the task sequence based on the compartment space and robot motion path, significantly improving stacking efficiency and reducing ineffective motion. It also deeply analyzes cargo constraint relationships, constructs a stacking sequence relationship diagram, and marks conflict points. This allows for rapid identification and resolution of cargo stacking sequence issues in path planning, ensuring that cargo is stacked in the correct order, avoiding damage or stacking failure due to incorrect sequences, and ensuring the accuracy and safety of operations.

[0016] 2. The present invention simulates the actual operation process of the robot through simulation verification, monitors the motion trajectory, posture and joint data in real time, performs collision detection and motion performance evaluation, and can discover potential collision risks, unstable motion and other problems in path planning in advance, avoid equipment damage and cargo loss in actual operation, and adjust the path in a targeted manner according to the simulation results to form an optimization iterative mechanism to ensure that the final global path planning solution meets the actual operation needs, continuously improve the quality and reliability of path planning, and enhance the adaptability and operation ability of the loading and stacking robot in complex scenarios.

[0017] 3. Based on the cargo stacking priority and compartment space layout, the present invention finely decomposes the task into subtasks corresponding to the stacking of individual cargo, determines the precise stacking coordinates and reserves spacing, and optimizes the subtask sorting in combination with the compartment space and the robot's motion path. It comprehensively collects and associates constraint rules, stores them, constructs a stacking sequence relationship graph, compares the initial local path with the relationship graph, automatically marks conflict points, and adjusts the path or operation sequence. It calculates and accumulates the time cost elements of each segment according to the robot's motion trajectory, providing quantitative indicators for path optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is the overall flow chart of the present invention.

[0019] Figure 2 The flowchart of constructing the three-dimensional space model of the carriage in the present invention.

[0020] Figure 3 This is a schematic diagram of the motion parameters of the loading and palletizing robot in the present invention. DETAILED DESCRIPTION

[0021] The present invention will be further described below with reference to the accompanying drawings.

[0022] A path planning method for an embodied intelligent loading and palletizing robot based on segmented path optimization is proposed. The embodied intelligent loading and palletizing robot path planning system adopted includes a data acquisition module, a path generation module, a path optimization module and a controller; the data acquisition module includes a laser radar and a depth camera; the laser radar is used to obtain point cloud data of the carriage; the depth camera is used to capture depth and texture information; the path generation module is used to generate an initial global path for the loading and palletizing robot to transport goods; the path optimization module is used to optimize the initial global path to obtain a final global path; and the controller is used to control the loading and palletizing robot according to the final global path.

[0023] like Figure 1 As shown, the path planning method of the embodied intelligent loading and stacking robot includes the following steps: Step 1: Figure 2As shown, lidar and depth cameras are used to collect basic information about the car, including car size, car shape and obstacle positions; the lidar emits laser beams at high frequency to obtain point cloud data, and the depth camera captures depth and texture information through structured light. The two types of data are unified in the coordinate system through coordinate conversion technology and then fused. The three-dimensional spatial model of the car is then constructed with the help of a three-dimensional modeling algorithm to accurately restore the car size, shape and obstacle positions, providing a real and reliable spatial reference for subsequent path planning.

[0024] Step 2: Use barcodes, QR codes, or RFID (radio frequency identification) technology to quickly read basic cargo information, including its size, shape, and weight. Build a cargo information database containing a basic information table, a priority table, and a constraint relationship table. Store and index this information to ensure rapid retrieval and real-time updates, forming the cargo data foundation for route planning. Store this information in the database, create an index to accelerate data retrieval, and promptly update the database whenever cargo information changes.

[0025] In this embodiment, for goods with complex shapes, a 3D scanner is used to obtain supplementary information on the dimensions of the goods.

[0026] Step 3: Figure 3 As shown, the motion parameters of the loading and palletizing robot are obtained, including movement parameters (maximum movement speed, acceleration, etc.), rotation parameters (maximum rotation angle, speed, etc.), and range parameters (horizontal / vertical working range, accessible space shape). The motion parameters define the loading and palletizing robot's movement capabilities and working boundaries within the vehicle compartment, ensuring that path planning conforms to the robot's actual performance and avoids exceeding its movement limits. Based on the loading and palletizing robot's motion parameters, robot constraints are constructed and stored in a constraint relationship table. Various constraints between goods are collected to construct cargo constraints, which are then associated with the corresponding goods in the cargo information database and stored in the constraint relationship table of the cargo information database.

[0027] Step 4: Construct the stacking order (stacking priority) of the goods based on the cargo constraints and store it in the constraint relationship table. Based on the basic information of the carriage, the basic information of the cargo, and the stacking priority, the cargo is matched with different spatial areas of the carriage. Within the selected spatial area, precise 3D stacking coordinates are determined for each cargo, taking into account the spacing between cargoes to ensure that the loading and stacking robots have sufficient operating space. The stacking process of each cargo is defined as an independent subtask. All subtasks are sorted according to the cargo stacking priority, with subtasks corresponding to cargo with higher priority placed first, ensuring that the loading and stacking robots perform stacking operations in the correct order.

[0028] Step 5. Allocate the goods to different loading and palletizing robots according to the cargo constraints; set the starting and ending points of the goods according to the stacking coordinates of the goods, and use the RRT algorithm to plan the path of each loading and palletizing robot to obtain the initial local path of each loading and palletizing robot for handling the goods. Combine the initial local paths of each loading and palletizing robot handling multiple goods to obtain the initial global path.

[0029] Step 6: Establish a time cost model and calculate the time cost of each initial local path; Using the initial global path as input, the time cost is integrated into a unified model. For each loading and palletizing robot, multiple initial global paths are obtained by adjusting the corresponding cargo and the order in which it handles the cargo. The time cost elements of each initial global path are calculated separately, namely, travel time, cargo grabbing / placing time, and waiting time. The total time cost of each initial global path is then summed to provide a quantitative metric for path optimization. The initial global path with the lowest time cost is selected as the optimized global path.

[0030] Step 7: Use simulated annealing algorithm to obtain the final global path Constructing the cost function .in, is the path length; is the obstacle avoidance cost (calculated based on the proximity of the path to the obstacle); The degree of compatibility between the path and the stacking method (e.g., whether the path reaches each stacking location in the planned stacking sequence); is the weight coefficient, which is adjusted according to factors such as actual logistics operation requirements, compartment environment characteristics, and robot performance to balance the impact of various factors on path optimization.

[0031] A new path is generated by perturbing the current path, and the Metropolis (Monte Carlo) criterion is used to determine whether to accept the new path. The acceptance probability formula is: ;in, is the cost function difference between the new path and the current path; is the current temperature. When , directly accept the new path; when When the probability Accept. Temperature According to the descent rate ( ) gradually decreases, i.e. ,As the temperature drops, the final global path of the robot carrying the goods is finally obtained.

[0032] Step 8: Motion Control ORCA (collision avoidance algorithm) is used to control the loading and stacking robot to move along the final global path. The specific process is as follows: Each palletizing robot is modeled as a circle, with its center at the robot's position and its radius as a preset safety buffer distance. The desired speed and direction are set for the palletizing robot; the desired direction coincides with the direction of the final global path. A palletizing robot whose distance from the current robot is less than a preset distance threshold is selected as its neighbor. A coordinate system is constructed with the current robot's position as the center, and the relative position and relative speed of each neighboring robot are obtained relative to the current robot. Based on the relative positions and speeds, the ORCA half-plane constraints between each robot and its neighbor are obtained. The speed and direction closest to the desired speed and direction in the ORCA half-plane constraints are selected as the final speed and direction of the corresponding palletizing robot. The palletizing robot is controlled to handle the cargo based on the speed and direction obtained at each moment.

[0033] Step 7: Simulation verification.

[0034] The final global path is input into the simulation system, allowing the loading and palletizing robot to move, grasp, and stack objects in the virtual environment according to this final global path. During the simulation, the robot's motion trajectory, posture changes, and joint motion data are recorded in real time. The simulation software's collision detection function monitors the robot's collision risk with the vehicle structure and stacked cargo, and also assesses whether its motion speed, acceleration, and smoothness meet the robot's constraints. Based on the simulation results, if issues such as collision, unstable motion, or excessive time cost are detected, the path planning strategy is adjusted accordingly, the global path is re-optimized, and simulation verification is performed again. Multiple iterations are performed until the path planning solution meets the requirements of efficient, safe, and accurate operation in the virtual environment. The verified path solution is finally applied to the actual loading and palletizing robot operation. During the simulation, the evaluation value f(n) of each node consists of the actual cost g(n) from the starting node to the current node and the estimated cost h(n) from the current node to the target node, i.e., f(n) = g(n) + h(n). Assume that the robot moves in the Cartesian coordinate system, and the step length of each movement is d. If it takes k steps from the starting node s to the current node n, then g(n)=k×d. For example, the robot moves from point ( x 1, y 1) Move to point ( x 2, y 2), the distance moved If you have passed k Such movement steps to reach the node n , then g(n) is the sum of the distances of these steps. Manhattan distance or Euclidean distance is usually used to estimate the distance from the current node n To the target nodet a. Manhattan distance: if the current node n The coordinates of x n , y n ), target node t The coordinates of x t , y t ),but ; b. Euclidean distance: .

Claims

1. A path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization, characterized by: The method includes: Obtain the basic information of the carriage and the cargo, and set the stacking priority of different cargoes based on the basic information of the carriage and the cargo; set the stacking coordinates of each cargo according to the stacking priority of the cargo; use the path planning algorithm to obtain the initial global path for each loading and stacking robot to transport the cargo; for each loading and stacking robot, obtain multiple initial global paths by adjusting the corresponding cargo it transports and the order in which the cargo is transported; calculate the total time cost of each initial global path separately, and select the initial global path with the lowest total time cost as the optimized global path; use the collision avoidance algorithm based on the optimized global path to obtain the movement direction and speed of the loading and stacking robot, so as to control the loading and stacking robot to transport the cargo.

2. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 1, characterized in that: After obtaining the optimized global path, the final global path is obtained through the simulated annealing algorithm.

3. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 2, characterized in that: The specific process of obtaining the final global path is as follows: construct a cost function; generate a new path by perturbing the current path, and determine whether to accept the new path based on the Monte Carlo criterion. The acceptance probability formula is: ;in, is the cost function difference between the new path and the current path; is the current temperature; when When , the new path is used as the current path; when When the probability Accept new path; temperature According to the descent rate Gradually reduce, that is ,As the temperature drops, the final global path of the robot carrying the goods is finally obtained.

4. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 3, characterized in that: The cost function The method to obtain is as follows: in, is the path length; Obstacle avoidance cost; The compatibility between the path and the stacking method; is the weight coefficient.

5. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 1, characterized in that: The method to obtain the movement direction and speed of the loading and stacking robot is as follows: Each loading and palletizing robot is modeled as a circle, with the center being the position of the loading and palletizing robot and the radius being the preset safety buffer distance; the expected speed and expected direction of the loading and palletizing robot are set; wherein the expected direction coincides with the direction of the final global path; the loading and palletizing robot whose distance from the current loading and palletizing robot is less than a preset distance threshold is selected as the adjacent loading and palletizing robot of the current loading and palletizing robot; the relative position and relative speed of each adjacent loading and palletizing robot and the current loading and palletizing robot are obtained; the collision avoidance algorithm half-plane constraint of each loading and palletizing robot and the adjacent loading and palletizing robot is obtained based on the relative position and relative speed; the speed and direction of the collision avoidance algorithm half-plane constraint closest to the expected speed and expected direction are selected as the final speed and movement direction of the corresponding loading and palletizing robot.

6. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 1, characterized in that: The basic information of the carriage includes the size, shape and location of obstacles; the basic information of the cargo includes the size, shape and weight of the cargo.

7. The path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization according to claim 1, characterized in that: The method for obtaining the basic information of the carriage is: respectively obtaining the point cloud data, depth and texture information of the loaded carriage, fusing the two types of data after unifying the coordinate system through coordinate conversion technology, and then constructing a three-dimensional spatial model of the carriage with the help of a three-dimensional modeling algorithm.

8. A path planning system for an embodied intelligent loading and palletizing robot based on segmented path optimization, characterized by: Used to execute the embodied intelligent loading and stacking robot path planning method based on segmented path optimization as described in claim 1; the embodied intelligent loading and stacking robot path planning system includes a data acquisition module, a path generation module, a path optimization module and a controller; the path generation module is used to generate an initial global path for the loading and stacking robot to stack goods; the controller is used to control the loading and stacking robot based on the acquired path.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: The memory stores a computer program; the processor executes a path planning method for an embodied intelligent loading and stacking robot based on segmented path optimization as described in any one of claims 1 to 7.

10. A readable storage medium storing a computer program; characterized in that: When the computer program is executed by a processor, it is used to implement the embodied intelligent loading and stacking robot path planning method based on segmented path optimization as described in any one of claims 1 to 7.

Images