Path planning method and system for industrial mobile robot

Abstract

The invention discloses a path planning method for an industrial mobile robot. The method comprises the following steps: establishing a coordinate system to describe a motion model of the robot; a wheel type odometer is used for obtaining pose data of the robot, an inertial measurement unit is used for obtaining speed data of the robot in the space, the speed data comprise linear acceleration and angular speed, and the pose data comprise the position and the pose of the robot; selecting an improved RBPF-based SLAM algorithm, and predicting and constructing a state estimation value of the robot and an unknown environment map according to a state estimation theory; based on the constructed environment map, acquiring nodes by adopting an improved A * bidirectional search algorithm, and sequentially connecting a plurality of nodes to obtain an overall path; according to the method, the optimal path from the starting point to the terminal point can be quickly found by adopting the improved A * bidirectional search algorithm on the constructed environment map. The algorithm is not only high in search speed, but also capable of avoiding falling into a local optimal solution and ensuring the global optimality of the path.

Description

Technical Field

[0001] The present invention relates to the technical field of path planning, and in particular to a path planning method and system for an industrial mobile robot. Background Art

[0002] In the field of industrial manufacturing, industrial mobile robots, as an automated equipment that integrates multiple technologies such as machinery, electronics, computers, information and automatic control, can be seamlessly integrated with workshop equipment, production lines and management information systems to achieve flexibility and intelligence. The main technologies involved in industrial mobile robots are intelligent perception, positioning and map construction, path planning and decision-making. These technologies are usually integrated into the robot's control system, enabling the robot to perform tasks effectively in complex industrial environments. For this reason, it is extremely important to study the intelligent technologies involved in industrial mobile robots.

[0003] Due to technical reasons, most researchers and technicians currently use data fusion technology for intelligent perception. The method used is based on particle filtering. Positioning and map construction are mainly based on visual SLAM technology. Path planning and decision-making are mainly based on Dijkstra, RRT and other algorithms. However, these technologies and algorithms have the following problems: (1) Poor generalization ability and limited applicable scenarios. (2) Easily affected by ambient light, resulting in inaccurate positioning. (3) Low search efficiency, and the route found is not the global optimal route. Summary of the invention

[0004] In view of the above-mentioned defects, the purpose of the present invention is to propose a path planning method and system for industrial mobile robots to solve the problems of low existing search efficiency and non-optimal paths.

[0005] To achieve this purpose, the present invention adopts the following technical solution: a path planning method for an industrial mobile robot, comprising the following steps:

[0006] Step S1: Establish a coordinate system to describe the motion model of the robot;

[0007] Step S2: using a wheel odometer to obtain the robot's posture data, and using an inertial measurement unit to obtain the robot's velocity data in space, wherein the velocity data includes linear acceleration and angular velocity, and the posture data includes the robot's position and posture;

[0008] Based on the extended Kalman filter algorithm, the posture data and velocity data are fused to obtain the robot's state estimation;

[0009] Step S3: Select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory;

[0010] Step S4: Based on the constructed environment map, an improved A* bidirectional search algorithm is used to obtain nodes, and multiple nodes are connected in sequence to obtain an overall path;

[0011] Step S5: using an artificial potential field method to obtain local paths between adjacent nodes;

[0012] Step S6: Build a hardware and software platform based on the ROS system, and set up a static environment and a dynamic environment to conduct physical verification experiments.

[0013] Preferably, in step S2, the steps of fusing the posture data and the velocity data based on the extended Kalman filter algorithm to obtain the state estimation of the robot are as follows:

[0014] Step S21: setting the state equation and observation equation;

[0015] Step S22: Initialize the state estimation value and covariance matrix in the Kalman filter;

[0016] Step S23: Input the posture data and velocity data for Kalman filtering.

[0017] Preferably, in step S3, the particle sampling method in the algorithm selects adaptive resampling, which is specifically:

[0018] The algorithm assumes that the state distribution is a single distribution, but uses an adaptive mixed proposal distribution to adjust the selection probability of each component according to the current velocity data and the state of the particle set.

[0019] Preferably, the specific steps of step S4 are as follows:

[0020] Step S41: Determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively;

[0021] Step S42: construct the first list and the second list, set the heuristic estimation value and initial cost of the start node and the target node to 0,

[0022] Entering the start node and the target node into the first list and the second list respectively;

[0023] Step S43: obtaining a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and storing the first best node in the first list;

[0024] Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list;

[0025] Step S44: Expand the first best node and the second best node to obtain multiple first best nodes and second best nodes, and store them in the first list and the second list respectively;

[0026] Step S45: after the node is expanded, determining whether there is an identical second best node in the first best node;

[0027] If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path;

[0028] If not present, repeat steps S43 to S44.

[0029] Preferably, the specific steps of step S5 are as follows:

[0030] Step S51: extracting turning points from the overall path;

[0031] Step S52: setting a distance parameter according to the distance between the turning points;

[0032] Step S53: selecting interval points evenly between connected turning points according to the distance parameter;

[0033] For each interval point, obtain the corresponding environmental data;

[0034] Step S54: combining the search strategy of interval points, planning the path points on the local path from one node to another adjacent node;

[0035] Step S55: using the least square method to perform line fitting on the path points;

[0036] Determine whether the fitted path reaches the adjacent node. If so, output the fitted path.

[0037] If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point;

[0038] If it enters, determine whether the repulsive force at the current position is greater than the attractive force;

[0039] If the repulsive force is greater than the attractive force, the repulsive force and attractive force function is updated, and step S54 is re-executed according to the new repulsive force and attractive force function; if the repulsive force is less than the attractive force, step S54 is re-executed;

[0040] If not, re-execute step S54.

[0041] A path planning system for an industrial mobile robot, using the path planning method for an industrial mobile robot, comprising:

[0042] A coordinate construction module is used to establish a coordinate system to describe the robot's motion model;

[0043] A data acquisition module is used to obtain the robot's posture data using a wheel odometer and obtain the robot's velocity data in space using an inertial measurement unit, wherein the velocity data includes linear acceleration and angular velocity, and the posture data includes the robot's position and posture; the posture data and velocity data are fused based on an extended Kalman filter algorithm to obtain a state estimation of the robot;

[0044] The map construction module is used to select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory;

[0045] The overall path module is used to obtain nodes using the improved A* bidirectional search algorithm, connect multiple nodes in sequence, and obtain the overall path;

[0046] The local path module is used to obtain the local path between adjacent nodes using the artificial potential field method;

[0047] The test module is used to build hardware and software platforms based on the ROS system, and to set up static and dynamic environments for physical verification experiments.

[0048] Preferably, the data acquisition module includes a setting unit, an initialization unit and a processing unit;

[0049] The setting unit is used to set the state equation and the observation equation;

[0050] The initialization unit is used to initialize the state estimation value and covariance matrix in the Kalman filter;

[0051] The processing unit is used to input the posture data and velocity data for Kalman filtering.

[0052] Preferably, the overall path module includes a determination unit, a storage unit, a configuration unit, an expansion unit, and a construction unit;

[0053] The determination unit is used to determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively;

[0054] The storage unit is used to construct the first list and the second list, set the heuristic estimation value and the initial cost of the start node and the target node to 0,

[0055] Entering the start node and the target node into the first list and the second list respectively;

[0056] The configuration unit is used to obtain a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and store the first best node in the first list;

[0057] Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list;

[0058] The expansion unit is used to expand the first best node and the second best node to obtain multiple first best nodes and second best nodes respectively, and store them in the first list and the second list respectively;

[0059] The construction unit is used to determine whether there is an identical second best node in the first best node after the node is expanded;

[0060] If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path.

[0061] Preferably, the local path module includes a first unit, a second unit, a third unit, a fourth unit and a fifth unit;

[0062] The first unit is used to extract turning points from the overall path;

[0063] The second unit is used to set the distance parameter according to the distance between the turning points;

[0064] The third unit is used to uniformly select interval points between connected turning points according to the distance parameter;

[0065] For each interval point, obtain the corresponding environmental data;

[0066] The fourth unit is used to plan the path points on the local path from one node to another adjacent node in combination with the search strategy of the interval points;

[0067] The fifth unit is used for performing route fitting on the path points using the least square method;

[0068] Determine whether the fitted path reaches the adjacent node. If so, output the fitted path.

[0069] If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point;

[0070] If it enters, determine whether the repulsive force at the current position is greater than the attractive force;

[0071] If the repulsive force is greater than the attractive force, the repulsive force and attractive force functions are updated, and the fourth unit is called again according to the new repulsive force and attractive force functions; if the repulsive force is less than the attractive force, the fourth unit is called again;

[0072] If not entered, the fourth unit is called again.

[0073] One of the above technical solutions has the following advantages or beneficial effects: In the present invention, by using an improved A* bidirectional search algorithm on a constructed environment map, the optimal path from the starting point to the end point can be quickly found. This algorithm not only has a fast search speed, but also can avoid falling into a local optimal solution, ensuring the global optimality of the path. At the same time, by connecting multiple nodes in sequence, a continuous and smooth overall path can be obtained, providing powerful guidance for the actual movement of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 is a flow chart of an embodiment of the method of the present invention.

[0075] Figure 2 It is a schematic diagram of the structure of an embodiment of the system of the present invention. DETAILED DESCRIPTION

[0076] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0077] In the description of the embodiments of the present invention, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0078] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0079] like Figures 1-2As shown, a path planning method for an industrial mobile robot includes the following steps:

[0080] Step S1: Establish a coordinate system to describe the motion model of the robot;

[0081] By accurately establishing the coordinate system, the robot's motion state in two-dimensional or three-dimensional space can be clearly described, providing a solid theoretical basis for subsequent data processing and path planning. This step ensures the accuracy and predictability of the robot's motion model, laying a solid foundation for the entire navigation process.

[0082] Step S2: using a wheel odometer to obtain the robot's posture data, and using an inertial measurement unit to obtain the robot's velocity data in space, wherein the velocity data includes linear acceleration and angular velocity, and the posture data includes the robot's position and posture;

[0083] Based on the extended Kalman filter algorithm, the posture data and velocity data are fused to obtain the robot's state estimation;

[0084] Using a variety of sensors such as wheel odometers and inertial measurement units (IMUs), the robot's position and velocity data can be obtained in real time. By fusing these data with the extended Kalman filter algorithm, a more accurate and reliable robot state estimation can be obtained. This data fusion method not only improves the accuracy of state estimation, but also enhances the robustness and stability of the system.

[0085] Step S3: Select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory;

[0086] Step S4: Based on the constructed environment map, an improved A* bidirectional search algorithm is used to obtain nodes, and multiple nodes are connected in sequence to obtain an overall path;

[0087] Step S5: using an artificial potential field method to obtain local paths between adjacent nodes;

[0088] Step S6: Build a hardware and software platform based on the ROS system, and set up a static environment and a dynamic environment to conduct physical verification experiments.

[0089] In the present invention, by using the improved A* bidirectional search algorithm on the constructed environment map, the optimal path from the starting point to the end point can be quickly found. This algorithm not only has a fast search speed, but also can avoid falling into the local optimal solution, ensuring the global optimality of the path. At the same time, by connecting multiple nodes in sequence, a continuous and smooth overall path can be obtained, providing powerful guidance for the actual movement of the robot.

[0090] Preferably, in step S2, the steps of fusing the posture data and the velocity data based on the extended Kalman filter algorithm to obtain the state estimation of the robot are as follows:

[0091] Step S21: setting the state equation and observation equation;

[0092] Step S22: Initialize the state estimation value and covariance matrix in the Kalman filter;

[0093] This step can be based on some prior knowledge or assumptions. The initial position can be assumed to be the origin, the velocity is 0, and a larger covariance matrix is ​​given to reflect this uncertainty.

[0094] Step S23: Input the posture data and velocity data for Kalman filtering.

[0095] In the present invention, Kalman filtering includes two steps of prediction and update:

[0096] Prediction steps:

[0097] Predicted state equation: x k =f(x k-1 ,u k-1 ,0), f(x) is the Jacobian matrix of the motion model of the odometer, x k-1 is the state estimate at time k-1, u k-1 is the speed data at time k-1;

[0098] Forecast estimate covariance matrix:

[0099] Update steps:

[0100] Calculate the Kalman gain:

[0101] Update the posterior state estimate: x kk =x k +K k (Z k -h(x k ,0));

[0102] Update the covariance matrix: P kk =(1-K k H k ) k ;

[0103] Among them A k is the transformation matrix at time k, P k-1 is the covariance of the estimation error at time k-1, W k is the system noise at time k, H k is the matrix that converts the state into measurement at time K, Vk is the observation noise at time k, x k is the state estimate at time k, Z k represents the actual measured value at time k, h(x k ,0) represents the state difference equation of the system.

[0104] Preferably, in step S3, the particle sampling method in the algorithm selects adaptive resampling, which is specifically:

[0105] The algorithm assumes that the state distribution is a single distribution, but uses an adaptive mixed proposal distribution to adjust the selection probability of each component according to the current velocity data and the state of the particle set.

[0106] Step A1: Generate a set of initial particles, each particle represents a possible robot pose and map feature set.

[0107] Step A2: For each particle, generate a new particle using the adaptive mixed proposal distribution and update its pose and map features.

[0108] Step A3: Calculate the weight of each particle based on the current velocity data.

[0109] Step A4: Determine whether to perform resampling based on the variance of the particle weight.

[0110] Step A5: Update the selection probability of the adaptive hybrid proposal distribution according to the current observation data and the state of the particle set.

[0111] Repeat steps A2-5 until the preset number of iterations is reached.

[0112] The present invention can reduce unnecessary resampling operations and maintain the diversity of particles through adaptive resampling. At the same time, it can more accurately describe complex state changes and improve the accuracy of the algorithm. Adaptive hybrid proposal distribution is used to adjust the selection probability of each component according to the current observation data and the state of the particle set. This can better adapt to the changes and uncertainties of the observation data and improve the sampling efficiency and estimation accuracy of the particle filter. The above improvements improve the positioning capability and search efficiency and can adapt to more complex scenarios.

[0113] Preferably, the specific steps of step S4 are as follows:

[0114] Step S41: Determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively;

[0115] Step S42: construct the first list and the second list, set the heuristic estimation value and initial cost of the start node and the target node to 0,

[0116] Entering the start node and the target node into the first list and the second list respectively;

[0117] Step S43: obtaining a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and storing the first best node in the first list;

[0118] Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list;

[0119] Step S44: Expand the first best node and the second best node to obtain multiple first best nodes and second best nodes, and store them in the first list and the second list respectively;

[0120] The node expansion method is the same as step S43, which is to select the first best node as the midpoint and find its adjacent nodes with the best heuristic estimation value and cost. Of course, the nodes stored in the first list are not included in the expansion range. When the second best node is expanded, it is also expanded in the above manner.

[0121] Step S45: after the node is expanded, determining whether there is an identical second best node in the first best node;

[0122] If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path;

[0123] If not present, repeat steps S43 to S44.

[0124] In the present invention, the first best node or the second best node is selected as the midpoint, and the adjacent node with the best heuristic estimation value and cost is searched for expansion. This expansion method not only considers the optimality of the current node, but also ensures the continuity of the path. During the search process, as new nodes are added and the list is updated, the algorithm can dynamically adjust the search direction to adapt to changes in the environment.

[0125] Preferably, the specific steps of step S5 are as follows:

[0126] Step S51: Extract turning points from the overall path; turning points are points where the path direction changes significantly, important nodes on the path (such as key points around obstacles), or user-defined key points.

[0127] Step S52: according to the distance between the turning points, set the distance parameter; the distance parameter determines how many interval points can be selected between two turning points. Specifically, the larger the distance between the turning points, the larger the distance parameter, wherein the distance parameter is the distance between adjacent interval points.

[0128] Step S53: selecting interval points evenly between connected turning points according to the distance parameter;

[0129] For each interval point, obtain the corresponding environmental data (such as obstacle location, terrain information, etc.);

[0130] Step S54: combining the search strategy of interval points (such as A* algorithm, Dijkstra algorithm, etc.), planning the path points on the local path from one node to another adjacent node;

[0131] Step S55: using the least square method to perform line fitting on the path points;

[0132] Determine whether the fitted path reaches the adjacent node. If so, output the fitted path.

[0133] If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point;

[0134] If it enters, determine whether the repulsive force at the current position is greater than the attractive force;

[0135] If the repulsive force is greater than the attractive force, the repulsive force and attractive force function is updated, and step S54 is re-executed according to the new repulsive force and attractive force function; if the repulsive force is less than the attractive force, step S54 is re-executed;

[0136] The initial repulsion function and attraction function are artificially set. When the calculated repulsion is greater than the attraction, it means that the current position may be too close to an obstacle. In the present invention, the magnitude of the repulsion and attraction functions will be adjusted to better adapt to the current environment. Re-execute step S54 to adjust the path.

[0137] If not, re-execute step S54.

[0138] In the present invention, an interval point search strategy and a range judgment strategy are added to achieve the purpose of optimization, and a least square method is added to the optimization potential field method to obtain a continuous and smooth path to meet the driving requirements of the robot.

[0139] A path planning system for an industrial mobile robot, using the path planning method for an industrial mobile robot, comprising:

[0140] A coordinate construction module is used to establish a coordinate system to describe the robot's motion model;

[0141] A data acquisition module is used to obtain the robot's posture data using a wheel odometer and obtain the robot's velocity data in space using an inertial measurement unit, wherein the velocity data includes linear acceleration and angular velocity, and the posture data includes the robot's position and posture; the posture data and velocity data are fused based on an extended Kalman filter algorithm to obtain a state estimation of the robot;

[0142] The map construction module is used to select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory;

[0143] The overall path module is used to obtain nodes using the improved A* bidirectional search algorithm, connect multiple nodes in sequence, and obtain the overall path;

[0144] The local path module is used to obtain the local path between adjacent nodes using the artificial potential field method;

[0145] The test module is used to build hardware and software platforms based on the ROS system, and to set up static and dynamic environments for physical verification experiments.

[0146] Preferably, the data acquisition module includes a setting unit, an initialization unit and a processing unit;

[0147] The setting unit is used to set the state equation and the observation equation;

[0148] The initialization unit is used to initialize the state estimation value and covariance matrix in the Kalman filter;

[0149] The processing unit is used to input the posture data and velocity data for Kalman filtering.

[0150] Preferably, the overall path module includes a determination unit, a storage unit, a configuration unit, an expansion unit, and a construction unit;

[0151] The determination unit is used to determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively;

[0152] The storage unit is used to construct the first list and the second list, set the heuristic estimation value and the initial cost of the start node and the target node to 0,

[0153] Entering the start node and the target node into the first list and the second list respectively;

[0154] The configuration unit is used to obtain a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and store the first best node in the first list;

[0155] Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list;

[0156] The expansion unit is used to expand the first best node and the second best node to obtain multiple first best nodes and second best nodes respectively, and store them in the first list and the second list respectively;

[0157] The construction unit is used to determine whether there is an identical second best node in the first best node after the node is expanded;

[0158] If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path.

[0159] Preferably, the local path module includes a first unit, a second unit, a third unit, a fourth unit and a fifth unit;

[0160] The first unit is used to extract turning points from the overall path;

[0161] The second unit is used to set the distance parameter according to the distance between the turning points;

[0162] The third unit is used to uniformly select interval points between connected turning points according to the distance parameter;

[0163] For each interval point, obtain the corresponding environmental data;

[0164] The fourth unit is used to plan the path points on the local path from one node to another adjacent node in combination with the search strategy of the interval points;

[0165] The fifth unit is used for performing route fitting on the path points using the least square method;

[0166] Determine whether the fitted path reaches the adjacent node. If so, output the fitted path.

[0167] If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point;

[0168] If it enters, determine whether the repulsive force at the current position is greater than the attractive force;

[0169] If the repulsive force is greater than the attractive force, the repulsive force and attractive force functions are updated, and the fourth unit is called again according to the new repulsive force and attractive force functions; if the repulsive force is less than the attractive force, the fourth unit is called again;

[0170] If not entered, the fourth unit is called again.

[0171] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0172] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A path planning method for an industrial mobile robot, characterized in that: The steps include: Step S1: Establish a coordinate system to describe the motion model of the robot; Step S2: using a wheel odometer to obtain the robot's posture data, and using an inertial measurement unit to obtain the robot's velocity data in space, wherein the velocity data includes linear acceleration and angular velocity, and the posture data includes the robot's position and posture; Based on the extended Kalman filter algorithm, the posture data and velocity data are fused to obtain the robot's state estimation; Step S3: Select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory; Step S4: Based on the constructed environment map, an improved A* bidirectional search algorithm is used to obtain nodes, and multiple nodes are connected in sequence to obtain an overall path; Step S5: using an artificial potential field method to obtain local paths between adjacent nodes; Step S6: Build a hardware and software platform based on the ROS system, and set up a static environment and a dynamic environment to conduct physical verification experiments.

2. A path planning method for an industrial mobile robot according to claim 1, characterized in that: In step S2, the posture data and velocity data are fused based on the extended Kalman filter algorithm to obtain the state estimation of the robot as follows: Step S21: setting the state equation and observation equation; Step S22: Initialize the state estimation value and covariance matrix in the Kalman filter; Step S23: Input the posture data and velocity data for Kalman filtering.

3. A path planning method for an industrial mobile robot according to claim 1, characterized in that: In step S3, the particle sampling method in the algorithm selects adaptive resampling, which is specifically: The algorithm assumes that the state distribution is a single distribution, but uses an adaptive mixed proposal distribution to adjust the selection probability of each component according to the current velocity data and the state of the particle set.

4. A path planning method for an industrial mobile robot according to claim 1, characterized in that: The specific steps of step S4 are as follows: Step S41: Determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively; Step S42: construct the first list and the second list, set the heuristic estimation value and initial cost of the start node and the target node to 0, Entering the start node and the target node into the first list and the second list respectively; Step S43: obtaining a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and storing the first best node in the first list; Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list; Step S44: Expand the first best node and the second best node to obtain multiple first best nodes and second best nodes, and store them in the first list and the second list respectively; Step S45: after the node is expanded, determine whether there is an identical second best node in the first best node; If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path; If not present, repeat steps S43 to S44.

5. A path planning method for an industrial mobile robot according to claim 1, characterized in that: The specific steps of step S5 are as follows: Step S51: extracting turning points from the overall path; Step S52: setting a distance parameter according to the distance between the turning points; Step S53: selecting interval points evenly between connected turning points according to the distance parameter; For each interval point, obtain the corresponding environmental data; Step S54: planning the path points on the local path from one node to another adjacent node in combination with the search strategy of the interval points; Step S55: using the least square method to perform line fitting on the path points; Determine whether the fitted path reaches the adjacent node. If so, output the fitted path. If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point; If it enters, determine whether the repulsive force at the current position is greater than the attractive force; If the repulsive force is greater than the attractive force, the repulsive force and attractive force function is updated, and step S54 is re-executed according to the new repulsive force and attractive force function; if the repulsive force is less than the attractive force, step S54 is re-executed; If not, re-execute step S54.

6. A path planning system for an industrial mobile robot, using a path planning method for an industrial mobile robot as claimed in any one of claims 1 to 5, characterized in that: include: A coordinate construction module is used to establish a coordinate system to describe the robot's motion model; A data acquisition module, used to obtain the robot's position and posture data using a wheel odometer, and to obtain the robot's velocity data in space using an inertial measurement unit, wherein the velocity data includes linear acceleration and angular velocity, and the position and posture data includes the robot's position and posture; Based on the extended Kalman filter algorithm, the posture data and velocity data are fused to obtain the robot's state estimation; The map construction module is used to select the SLAM algorithm based on the improved RBPF, and predict and construct the robot's state estimation and unknown environment map according to the state estimation theory; The overall path module is used to obtain nodes using the improved A* bidirectional search algorithm, connect multiple nodes in sequence, and obtain the overall path; The local path module is used to obtain the local path between adjacent nodes using the artificial potential field method; The test module is used to build hardware and software platforms based on the ROS system, and to set up static and dynamic environments for physical verification experiments.

7. A path planning system for an industrial mobile robot according to claim 6, characterized in that: The data acquisition module includes a setting unit, an initialization unit and a processing unit; The setting unit is used to set the state equation and the observation equation; The initialization unit is used to initialize the state estimation value and covariance matrix in the Kalman filter; The processing unit is used to input the posture data and velocity data for Kalman filtering.

8. A path planning system for an industrial mobile robot according to claim 6, characterized in that: The overall path module includes a determination unit, a storage unit, a configuration unit, an expansion unit, and a construction unit; The determination unit is used to determine the starting point and the end point of the robot, and set the starting point and the end point as the starting node and the target node respectively; The storage unit is used to construct the first list and the second list, set the heuristic estimation value and the initial cost of the start node and the target node to 0, Entering the start node and the target node into the first list and the second list respectively; The configuration unit is used to obtain a node adjacent to the start node with the best heuristic estimation value and cost as the first best node, and store the first best node in the first list; Obtaining a node adjacent to the target node with the best heuristic estimation value and cost as the second best node, and storing the second best node in a second list; The expansion unit is used to expand the first best node and the second best node to obtain multiple first best nodes and second best nodes respectively, and store them in the first list and the second list respectively; The construction unit is used to determine whether there is an identical second best node in the first best node after the node is expanded; If it exists, the node is marked as an encounter node, and starting from the encounter node, the parent nodes in the starting direction and the target direction are tracked respectively until the starting node and the target node are reached to obtain the overall path.

9. A path planning system for an industrial mobile robot according to claim 6, characterized in that: The local path module includes a first unit, a second unit, a third unit, a fourth unit and a fifth unit; The first unit is used to extract turning points from the overall path; The second unit is used to set the distance parameter according to the distance between the turning points; The third unit is used to uniformly select interval points between connected turning points according to the distance parameter; For each interval point, obtain the corresponding environmental data; The fourth unit is used to plan the path points on the local path from one node to another adjacent node in combination with the search strategy of the interval points; The fifth unit is used for performing route fitting on the path points using the least square method; Determine whether the fitted path reaches the adjacent node. If so, output the fitted path. If not, calculate the distance between the current position and the next interval point to determine whether it is within the core distance of the interval point; If it enters, determine whether the repulsive force at the current position is greater than the attractive force; If the repulsive force is greater than the attractive force, the repulsive force and attractive force functions are updated, and the fourth unit is called again according to the new repulsive force and attractive force functions; if the repulsive force is less than the attractive force, the fourth unit is called again; If not entered, the fourth unit is called again.

Images