A method and system for intelligent navigation planning of mobile robots

Abstract

This invention relates to the field of autonomous robot navigation technology, specifically to an intelligent navigation planning method and system for a mobile robot. The method includes: generating a global path based on a global path planning algorithm; optimizing and adjusting the global path using a local path optimization algorithm in conjunction with real-time environmental information; and adjusting the global path in real-time to avoid dynamic obstacles, thereby obtaining an optimized path. This addresses the problems in existing technologies where path planning schemes have poor adaptability in dynamic environments, struggle to meet real-time requirements, and cannot adaptively adjust strategies to plan the optimal path. The system includes an environmental perception module, a global path planning module, a local path optimization module, and a robot control module. Information exchange between these modules is achieved through standardized interfaces, significantly improving path planning efficiency and navigation performance in dynamic environments.

Description

Technical Field

[0001] This invention belongs to the field of robot autonomous navigation technology, specifically relating to an intelligent navigation planning method and system for a mobile robot. Background Technology

[0002] With the rapid development of artificial intelligence and autonomous robot navigation technologies, mobile robots have been widely used in various fields such as logistics, security, and healthcare. Currently, mobile robot path planning technologies in the industry, both domestically and internationally, are mainly divided into two categories: rule-based path planning methods and machine learning-based path planning methods.

[0003] Rule-based path planning methods typically employ classic algorithms such as Dijkstra's algorithm and A* algorithm. These algorithms aim for global optimization and plan paths based on a pre-known static environment map, which can guarantee the feasibility and optimality of the planned path to a certain extent. However, these methods are poorly adaptable to dynamic environments. When dynamic obstacles appear or the path changes, the path often needs to be replanned, resulting in long computation times. Furthermore, they cannot handle high-dimensional spaces. When robots need to navigate in high-dimensional state spaces (such as complex 3D scenes), the computational complexity of these algorithms increases rapidly, making it difficult to meet real-time requirements. Simultaneously, these algorithms lack intelligence; traditional algorithms rely on fixed rules and global information and cannot adaptively adjust strategies or optimize paths.

[0004] At the same time, machine learning-based methods also have limitations. With the development of reinforcement learning techniques, deep strong learning is widely used in robot path planning, such as Deep Q-learning (DQN) and Proximal Policy Optimization (PPO) algorithms. These methods enable robots to complete navigation tasks in unknown or dynamic environments through exploration and learning of the environment. However, this method also has the following drawbacks:

[0005] 1. Slow convergence speed: Existing reinforcement learning algorithms often require a lot of training time to converge when dealing with complex environments, resulting in high development costs.

[0006] 2. Low exploration efficiency: Since reinforcement learning relies on random sampling, it may lead to repeated exploration of low-value areas, prolonging training time.

[0007] 3. Insufficient optimality of the planned path: Reinforcement learning algorithms often cannot match the global optimality of traditional algorithms and are prone to producing suboptimal paths.

[0008] Based on the above analysis, existing technologies are insufficient to simultaneously meet the requirements of dynamic environment adaptability, real-time performance, high-dimensional environment planning capabilities, and global path optimization. Therefore, this invention is proposed. Summary of the Invention

[0009] The purpose of this invention is to provide an intelligent navigation planning method and system for mobile robots, thereby solving the problems of poor adaptability of existing path planning schemes in dynamic environments, difficulty in meeting real-time requirements, and inability to adaptively adjust strategies to plan the optimal path.

[0010] The objective of this invention is achieved through the following technical solution:

[0011] A method for intelligent navigation planning of a mobile robot, comprising:

[0012] Real-time collection of environmental information, including static map information and real-time environmental information;

[0013] Based on the static map information, a global path planning algorithm is used to plan a global path;

[0014] Based on the real-time environmental information, a local path optimization algorithm is used to optimize and adjust the global path, adjusting the global path in real time to avoid dynamic obstacles and obtain an optimized path.

[0015] Control commands are generated based on the optimized path, and the robot is driven to move along the optimized path by the control commands.

[0016] The global path optimization adjustment uses the Soft Actor / Critic algorithm, and the process includes:

[0017] Input the global path and the real-time environmental information contained in the environmental information;

[0018] Define a state space S and an action space A, and determine the information contained in the state space S and the action space A based on the environmental information;

[0019] Based on the information contained in the state space S and action space A, a dynamic reward function is determined; the dynamic reward function is used to update the relationship between the robot and dynamic obstacles; wherein, the dynamic reward function includes: rewarding behaviors that approach the target point, penalizing behaviors that approach dynamic obstacles, and penalizing behaviors that deviate from the global path; the path planning strategy is updated based on the above dynamic reward function and combined with the soft actor critic algorithm; the global path is then locally optimized based on the updated path planning strategy to obtain the optimized path.

[0020] The motion space A includes the robot's selectable motion direction and speed; the state space S includes the robot's current position, speed, and local environmental features.

[0021] The dynamic reward function is determined based on the environmental information, and specifically, the dynamic reward function is calculated based on the following formula:

[0022] R(s, a) = α1·d target +α2·d obstacle +α3·d A*

[0023] Where, d target Let d be the distance from the robot to the target point. obstacle d is the distance from the robot to the nearest obstacle. A* The offset distance from the robot to the global path is α1, α2, and α3, which are weight parameters. The weight parameters are set and adjusted according to the path planning requirements.

[0024] The update of the path planning strategy includes strategy network optimization and value function optimization;

[0025] Policy network optimization: Maximize the objective function, which is expressed as: J(π) = E st，at [R(s, a)];

[0026] Where E is the expectation function, st is the current environmental state, at is the action selected according to policy π in state st, and R(s, a) is the dynamic reward function.

[0027] Value function optimization: Minimize the Bellman residual and update the parameters of the policy network.

[0028] The real-time environmental information includes: road condition information, dynamic obstacle information, and robot motion status.

[0029] The global path planning uses the A* algorithm.

[0030] The static map information includes: the robot's current position, the target position, and the global environment map; the control commands include: navigation commands, obstacle avoidance priority commands, and energy-saving mode commands.

[0031] The intelligent navigation planning method further includes a navigation experience optimization strategy, which includes:

[0032] After each task is completed, the robot stores the state, actions, and reward information from the path planning process into a shared experience pool, gradually optimizing the policy network of the soft actor critic algorithm.

[0033] An intelligent navigation planning system for a mobile robot includes:

[0034] Environmental perception module: used to collect environmental information in real time, including static map information and real-time environmental information;

[0035] Global path planning module: Used to receive static map information sent by the environment perception module and generate global paths through the global path planning algorithm;

[0036] Local path optimization module: It is used to receive real-time environmental information sent by the environmental perception module, and use the local path optimization algorithm to optimize and adjust the global path. The global path is adjusted in real time to avoid dynamic obstacles and obtain an optimized path.

[0037] Robot control module: Used to receive the optimized path sent by the local path optimization module and the real-time environmental information sent by the environmental perception module, generate control commands based on the optimized path, and drive the robot to move along the optimized path through the control commands.

[0038] The environmental perception module includes a lidar and a depth camera, used to collect information on static obstacles around the robot, dynamic obstacles, and the robot's own state.

[0039] Compared with existing technologies, the intelligent navigation planning method and system for mobile robots provided by this invention, in specific applications, combines the A* algorithm and the SAC algorithm to achieve global path planning and real-time optimization of the dynamic environment, effectively improving the navigation accuracy and obstacle avoidance capabilities of mobile robots in complex environments. Furthermore, by adjusting path planning through reinforcement learning strategies, it effectively overcomes the limitations of traditional path planning methods in dynamic environments, improving the robot's flexibility and navigation efficiency in practical applications. Attached Figure Description

[0040] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A schematic diagram of the intelligent navigation planning method provided in an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the architecture of an intelligent navigation planning system provided in an embodiment of the present invention. Detailed Implementation

[0043] The technical solution of the present invention will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are not all embodiments of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0044] It should be noted that, unless otherwise specifically stated, the relative arrangement and numerical expressions of the components and steps described in these embodiments should not be construed as limiting the scope of the invention.

[0045] The following description of exemplary embodiments is merely illustrative and is not intended to limit the invention or its application or use in any way. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail herein, but where applicable, such techniques, methods, and apparatus should be considered part of this specification.

[0046] Explanation of technical terms:

[0047] A* algorithm: short for A-Star algorithm;

[0048] SAC algorithm: short for Soft Actor-Critic algorithm.

[0049] Example 1

[0050] See Figure 1 The present invention provides a schematic diagram of an intelligent navigation planning method for a mobile robot, the specific implementation of which may include the following steps:

[0051] A1. Real-time collection of environmental information

[0052] Environmental information is collected in real time using sensor devices. This environmental information includes static map information and real-time environmental information. The static map information includes the robot's current position, target position, and global environment map. The real-time environmental information includes road condition information, dynamic obstacle information, and the robot's motion state. The specific steps for real-time environmental information collection are as follows:

[0053] A11. Collect static map information and real-time environmental information of the surrounding environment in real time through sensors;

[0054] A12. Combine dynamic obstacle information and robot motion status to update real-time environmental information in real time to ensure the real-time nature and accuracy of environmental data when planning the path;

[0055] A13. By integrating static map information and real-time environmental information, multi-level environmental perception data is generated to support global planning and local optimization.

[0056] A2. Global Path Planning

[0057] Based on the static map information, the A* algorithm is used to plan a global path. As a classic heuristic global path planning method, the A* algorithm provides a global path based on the static map information. Based on heuristic search, the A* algorithm can quickly find the globally optimal path. Through the planning using the A* algorithm, it is ensured that the robot can reach the target point along the shortest path, without considering dynamic obstacles. The specific operation is as follows:

[0058] A21. Construct a gridded map based on the static map information, and define the traversability and cost function of each unit included in the gridded map;

[0059] A22. Calculate the cost value f(n) = g(n) + h(n) for each node in the gridded map using a heuristic search strategy;

[0060] Where g(n) is the actual cost from the starting point to node n; h(n) is the estimated cost from node n to the target point;

[0061] A23. Gradually expand from the starting point to the target point to generate a globally optimal path, and mark the path nodes to form an initial path guide, which serves as the global path obtained from the planning.

[0062] A24. Input the planned global path results into the local optimization module of the next stage as a reference.

[0063] A3. Local Path Optimization

[0064] By combining the environmental information, a local path optimization algorithm is used to locally optimize and adjust the global path, adjusting the path in real time to avoid dynamic obstacles, thus obtaining an optimized path. The global path optimization and adjustment can employ the SAC algorithm; its specific implementation steps are as follows:

[0065] A31. Input the global path and real-time environment information planned by the A* algorithm;

[0066] A32. Define the state space S and action space A for reinforcement learning, where:

[0067] The state space S includes: the robot's current position, velocity, and local environmental features;

[0068] The motion space A includes: the robot's selectable motion direction and speed;

[0069] The information included in the state space S and action space A is determined based on the real-time environment information.

[0070] A33. Based on the information contained in the state space S and action space A, a dynamic reward function is determined; the dynamic reward function is used to update the relationship between the robot and dynamic obstacles; wherein, the dynamic reward function includes: rewarding behaviors that approach the target point, penalizing behaviors that approach dynamic obstacles, and penalizing behaviors that deviate from the global path. The formula for the dynamic reward function is:

[0071] R(s, a) = α1·d target +α2·d obstacle +α3·d A*

[0072] Where, d target Let d be the distance from the robot to the target point. obstacle d is the distance from the robot to the nearest obstacle. A* α1, α2, and α3 represent the offset distance from the robot to the global path, and α3 are weight parameters. These weight parameters can be set and adjusted according to path planning requirements.

[0073] A34. Update the path planning strategy based on the dynamic reward function and the SAC algorithm; perform local optimization on the global path based on the updated path planning strategy to obtain the optimized path; the update of the path planning strategy includes policy network optimization and value function optimization; policy network optimization: maximize the objective function, whose expression is: J(π)=E st，at [R(s, a)];

[0074] Where E is the expectation function, st is the current environmental state, at is the action selected according to policy π in state st, and R(s, a) is the dynamic reward function.

[0075] The optimization objective of the policy network is to learn a policy π(a|s), which is the probability distribution of the robot choosing different actions a in any state s. Through this policy, the robot can select the optimal action at each time step, thereby maximizing the cumulative reward of the entire path planning task. The optimization of the policy network directly determines the robot's real-time decision-making capability in local path planning, especially achieving efficient path optimization in dynamic environments.

[0076] Value function optimization: Minimize the Bellman residuals, the parameters of the policy network;

[0077] The goal of value function optimization is to learn an action value function Q(s, a), which is the expected cumulative reward that may be obtained in the future after choosing action a in state s. The value function provides guidance for the training of the policy network, helping the policy network to more clearly evaluate the merits of each action during the optimization process, avoiding blind exploration, thereby improving the learning efficiency and stability of the policy.

[0078] Through the synergistic effect of policy network optimization and value function optimization, it is possible to dynamically adapt to complex environments and guide robots to make efficient real-time decisions in local path planning.

[0079] A35. Generate new motion control commands for the robot based on the optimized path, so as to guide the robot to effectively avoid dynamic obstacles during movement.

[0080] The local path optimization process provided in step A3 above addresses the poor adaptability of the A* algorithm in dynamic environments through dynamic optimization, significantly improving the real-time performance and robustness of path planning. It combines the advantages of the A* and SAC algorithms to adjust the path in real time to adapt to complex and changing environments.

[0081] Furthermore, in the above processing, the A* algorithm is responsible for global path planning, quickly generating an initial path to provide global guidance; the SAC algorithm, on the other hand, achieves local optimization and dynamic obstacle avoidance through reinforcement learning. This approach leverages the A* algorithm to reduce the inefficiency of random exploration while simultaneously utilizing the SAC algorithm for real-time environmental perception and path optimization, further improving the response speed of path planning while ensuring the rationality of the global path. Therefore, the combination of the A* and SAC algorithms effectively overcomes the limitation of slow response in dynamic environments by traditional methods, thus meeting the real-time and adaptive navigation requirements of mobile robots.

[0082] A4. The robot executes control commands.

[0083] The system generates control commands based on the optimized path, and drives the robot to move along the optimized path using these commands. It also includes receiving real-time environmental information and generating emergency commands based on this information. The real-time environmental information includes road conditions, obstacle locations, and the robot's motion status. The road conditions include detailed information such as ground conditions and ground inclination. The specific implementation steps are as follows:

[0084] A41. By receiving key nodes and robot state information in the optimized path, the path points are converted into target pose points;

[0085] A42. Based on the target pose point and the robot kinematic model, calculate the control parameters of the robot's speed and direction, and then generate corresponding control commands.

[0086] A43. The real-time environmental information is converted into a series of pose target points. When the pose target points overlap with the optimized path pose target points, they are marked as dynamic obstacles. When the robot cannot avoid dynamic obstacles by relying solely on the optimized path, an emergency command is generated based on the real-time environmental information. The emergency command includes operation commands such as stop movement command and deceleration command.

[0087] A44. Send the control command to the underlying drive module to drive the robot to move along the optimized path. When an emergency command is received, stop the control command and execute the emergency command.

[0088] Furthermore, during the robot's movement, the deviation between the robot's position and path can be monitored in real time to ensure the accuracy of path planning.

[0089] A5. Navigation Experience Optimization Strategies

[0090] After each task is completed, the robot stores the state, actions, and reward information of the path planning strategy process into a shared experience pool, gradually optimizing the policy network of the SAC algorithm and improving the planning efficiency and dynamic response capability for future tasks.

[0091] The above processing procedure is applicable to various types of mobile robot tasks. By adjusting the algorithm weight parameters α1, α2, and α3, it can be adapted to different task requirements, such as efficient navigation, obstacle avoidance priority, or energy-saving mode, thereby improving the versatility and practicality of the algorithm and making it suitable for various industry applications.

[0092] In summary, this invention can meet the real-time path planning requirements in complex dynamic environments. Whether it's efficient navigation, obstacle avoidance priority, or energy-saving mode requirements, all can be achieved by adjusting the algorithm weight parameters. Through adaptive optimization using reinforcement learning and global path guidance, the efficiency, flexibility, and safety of path planning are significantly improved, providing a reliable guarantee for the application of mobile robots in dynamic environments.

[0093] Example 2

[0094] See Figure 2 , Figure 2 This is a schematic diagram of the structure of a mobile robot intelligent navigation system proposed in an embodiment of the present invention. The system employs a modular design and data interfaces to enable communication between modules, ensuring real-time information updates and accurate decision-making, thereby improving the stability and flexibility of the navigation system. Figure 2 As shown, the mobile robot intelligent navigation system may specifically include:

[0095] B1. Environmental Perception Module

[0096] The environmental perception module is used to collect environmental information in real time, including static map information and real-time environmental information, and sends this information to the global path planning module, the local path optimization module, and the robot control module. This module collects environmental data through various sensors (such as LiDAR, depth cameras, IMU sensors, etc.) to ensure data comprehensiveness and real-time performance. The environmental information includes information on static obstacles around the robot, dynamic obstacles, and the robot's own state. Specific functions include:

[0097] B11. Static Obstacle Information Acquisition: Generate a global static map through environmental scanning, including the location, size, and shape of obstacles;

[0098] B12. Dynamic obstacle information perception: Detects the position, speed, and direction of dynamic obstacles to support real-time path optimization;

[0099] B13. Self-state information monitoring: Record the robot's current position, speed, and direction to ensure that the path planning is consistent with the robot's actual state.

[0100] The static map information generated by the environmental perception module is sent to the global path planning module through a data interface for planning the global path; the real-time environmental information generated by the module is sent to the local path optimization module and the robot control module for real-time path optimization and navigation tasks.

[0101] B2, Global Path Planning Module

[0102] The global path planning module receives environmental information sent by the environmental perception module and generates a global path using a global path planning algorithm. The global path planning module uses the A* algorithm to generate the global path. Specific functions include:

[0103] B21. Construct a global static grid map based on the environmental information: perform grid processing on the environment and define passable areas and obstacle areas;

[0104] B22. Using a heuristic search strategy to plan a path: The A* algorithm finds an optimal path from the starting point to the target point by evaluating the cost function f(n) = g(n) + h(n);

[0105] B23. Output global path: Send the global path as a reference path to the local path optimization module to provide a basis for subsequent dynamic adjustments.

[0106] The global path planning module features high computational efficiency and intuitive planning results, providing global navigation direction guidance for the robot. However, it does not handle interference from dynamic obstacles and needs to be combined with the local path optimization module to complete dynamic obstacle avoidance.

[0107] B3. Local Path Optimization Module

[0108] The local path optimization module receives environmental information sent by the environmental perception module and performs local optimization and adjustment on the global path using a local path optimization algorithm. It adjusts the global path in real time to avoid dynamic obstacles and obtains an optimized path. The local path optimization module uses the SAC algorithm to perform local optimization and adjustment on the global path.

[0109] Specific functions include:

[0110] B31, Environmental Information Receiving Unit: Used to receive the global path and the real-time environmental information contained in the environmental information;

[0111] B32. Spatial definition unit: used to define the state space S and the action space A, and to determine the information contained in the state space S and the action space A based on the environmental information;

[0112] B33. Dynamic Reward Function Generation Unit: Used to dynamically establish a dynamic reward function based on the information contained in the state space S and action space A; the dynamic reward function is used to update the relationship between the robot and dynamic obstacles; wherein, the dynamic reward function includes: rewarding behavior that approaches the target point, penalizing behavior that approaches dynamic obstacles, and penalizing behavior that deviates from the global path; the dynamic reward function is designed considering the distance from the robot to the target point, the distance from the robot to the obstacle, and the degree of deviation of the robot from the global path, and its formula is as follows:

[0113] R(s, a) = α1·d target +α2·d obstacle +α3·d A*

[0114] Where, d target Let d be the distance from the robot to the target point. obstacle d is the distance from the robot to the nearest obstacle. A* α1, α2, and α3 represent the offset distance from the robot to the global path, and α3 are weight parameters. These weight parameters can be set and adjusted according to path planning requirements.

[0115] B34. Policy Update Unit: Used to learn a policy network that can guide the robot to make real-time decisions in complex environments, helping the robot to complete path planning and local optimization in dynamic environments; guiding the robot to select the optimal action at each time step, thereby maximizing the cumulative reward of the entire path planning task; defining the robot's real-time decision-making capability in local path planning, used for local path optimization in dynamic environments.

[0116] It also includes value function optimization, which is used to learn an action value function. The action value function provides guidance to the policy network and assists in the training of the policy network. The value function provides guidance to the policy network and helps the policy network to more clearly evaluate the merits of actions during the optimization process.

[0117] B35. Optimized Path Generation Unit: Based on the updated path planning strategy, the global path is locally optimized to obtain the optimized path.

[0118] By adjusting the weight parameters, the system can adapt to different task scenarios, enabling flexible strategy optimization. This module utilizes reinforcement learning to achieve dynamic obstacle avoidance and global path optimization, enhancing the system's dynamic adaptability and environmental responsiveness. Simultaneously, the introduction of the policy update unit allows for a more explicit evaluation of the merits of a particular action, avoiding blind exploration and improving the learning efficiency and stability of the strategy.

[0119] The aforementioned global path planning module and local path optimization module interact through a standardized interface, which supports real-time data updates and modular expansion. The global path planning module and local path optimization module can operate independently and support the expansion and upgrading of the environment perception module and control module to meet different task requirements and scenario changes.

[0120] B4. Robot Control Module

[0121] The robot control module receives the optimized path from the local path optimization module and real-time environmental information from the environment perception module. Based on the optimized path, it generates control commands to drive the robot to move along the optimized path. The real-time environmental information includes road condition information, obstacle locations, and the robot's motion status. The road condition information includes detailed information such as ground conditions and ground inclination. Specific functions include:

[0122] B41. Real-time environmental information analysis unit: used to convert real-time environmental information into pose target point information; when the pose target point is repeated with the optimized path pose target point, it is marked as a dynamic obstacle.

[0123] B42. Control command generation unit: used to calculate the control parameters of the robot's speed and direction based on the target pose point and the robot kinematic model, and then generate corresponding control commands;

[0124] B43. Emergency Command Generation Unit: Used to convert the real-time environmental information into a series of pose target points. When the pose target points are repeated with the optimized path pose target points, they are marked as dynamic obstacles. When the robot cannot avoid dynamic obstacles by relying solely on the optimized path, it generates emergency commands based on the real-time environmental information. The emergency commands include operation commands such as stop movement commands and deceleration commands.

[0125] B44. Drive Module Execution Unit: Used to send the control commands to the underlying drive module, drive the robot to move along the optimized path, and when an emergency command is received, stop the control commands and execute the emergency command.

[0126] B45. Navigation Experience Optimization Strategy Unit: After each task is completed, the robot stores the state, action, and reward information from the path planning strategy process into a shared experience pool, gradually optimizing the policy network of the SAC algorithm and improving the planning efficiency and dynamic response capability for future tasks.

[0127] This system adopts a modular design and uses standardized interfaces to achieve information exchange between modules, which has the following advantages:

[0128] 1. Combining global and local approaches: By combining the A* algorithm with the SAC algorithm, we can respond to dynamic environments in real time and make up for the limitations of a single algorithm.

[0129] 2. Independent modular design: Each module operates independently, which facilitates development, debugging, and later maintenance, and reduces system complexity;

[0130] 3. Flexible expansion and upgrade: Through modular design and standardized interfaces, new functional modules can be easily added or existing modules can be optimized to meet system expansion needs;

[0131] 4. Rapid deployment capability: The reinforcement learning module reduces training time, significantly shortens the development cycle, and supports rapid deployment and application;

[0132] 5. Real-time performance and safety: It can not only quickly adjust robot behavior in complex environments, but also make autonomous adjustments even when the path planning is not fully covered or when there are emergencies in the planned path;

[0133] 6. Efficient data processing: The system achieves an efficient closed loop of environmental perception, path planning, strategy optimization and task execution, ensuring real-time information updates and accurate path planning.

[0134] Example 3

[0135] This embodiment demonstrates the advantages of combining the A* algorithm and the SAC algorithm to achieve efficient and practical path planning and dynamic obstacle avoidance in a dynamic warehousing and logistics simulation environment. The following details the system construction, algorithm design, operation process, and experimental results analysis.

[0136] C1. System Construction and Initialization

[0137] C11, Hardware Equipment

[0138] C111 Mobile Robot: An autonomously navigational mobile robot with precise positioning and flexible movement capabilities. Its chassis integrates multiple degrees of freedom motion wheels, enabling omnidirectional movement. The built-in processor supports real-time algorithm calculations and path optimization.

[0139] C112, Sensor Devices: As the environmental perception module, it may include a lidar and a depth camera; the lidar provides 360° environmental scanning capability to detect the distance and position of surrounding obstacles, dynamic obstacle information and robot's own state information; the depth camera is used to capture the depth information of the environment, especially to support the identification and speed estimation of dynamic obstacles.

[0140] C113. Communication Equipment: Equipped with a high-speed wireless communication module, it can realize task sharing and collaborative path planning in multi-robot systems, while also supporting remote control and real-time monitoring.

[0141] C12, Task Scenarios

[0142] C121. Site Setup: The simulation environment simulates a 100-square-meter dynamic warehousing and logistics site, including fixed obstacles (such as shelves and pillars) and dynamic obstacles (such as other mobile robots or workers). Delivery points are randomly distributed, and the dynamic obstacles in the environment move at different speeds and directions, simulating the dynamic complexity of real-world warehousing scenarios.

[0143] C122. Task Requirements: The robot must start from the starting point, accurately avoid both fixed and dynamic obstacles in the area, and safely transport the goods to the target point. The task objective and path may change dynamically with the environment.

[0144] C13, System Construction and Initialization

[0145] C131, Global Path Planning Module: Used to construct a static map of the site, which is then gridded into several nodes. A global path is generated using the heuristic A* algorithm based on node cost values, providing a global path reference for the task.

[0146] C132, Local Path Optimization Module: Used to receive environmental information sent by the environmental perception module, use the SAC algorithm to optimize and adjust the global path, adjust the global path in real time to avoid dynamic obstacles, and obtain the optimized path;

[0147] C133. Robot Control Module: Used to receive the optimized path sent by the local path optimization module and the environmental information sent by the environmental perception module, generate control commands based on the optimized path, and drive the robot to move along the optimized path through the control commands.

[0148] C2. Operation Process

[0149] C21. Task Allocation

[0150] After the system starts, target points are assigned to the robot according to the task requirements. The robot plans the global path using the A* algorithm to obtain the initial path from the starting point to the target point, and then uses this path as a reference to pass it into the local path optimization module for optimization.

[0151] C22, Path Planning and Optimization

[0152] Global path planning: The robot completes initial navigation in a static environment based on the global path provided by the A* algorithm. This global path is then passed to the local optimization module as one of the constraints for subsequent adjustments.

[0153] Local path optimization: The SAC algorithm perceives real-time environmental information and identifies the type, location, and movement trend of obstacles.

[0154] The dynamic reward function dynamically adjusts the weight parameters: when obstacles are dense or the dynamic environment is complex, the obstacle avoidance weight is increased to prioritize the success rate of obstacle avoidance; when the path is relatively smooth, the path length optimization weight is increased to ensure driving efficiency.

[0155] SAC generates the optimal action instruction for the current time step through trial and error and learning, and updates the global path in real time.

[0156] C23. Task Execution and Dynamic Adjustment

[0157] The robot follows the optimized path to perform navigation tasks and uploads surrounding environmental data to the system's shared pool. The data in the shared pool supports real-time collaboration among other robots, avoiding task conflicts.

[0158] When obstacles or changes in the environment are detected that obstruct the path, the SAC algorithm re-optimizes the path based on the new environmental state to ensure that the task can be completed continuously and efficiently.

[0159] C24. Feedback and Learning

[0160] After each task is completed, the robot stores the state, actions, and reward information from the path planning process into a shared experience pool, gradually optimizing the policy network of the SAC algorithm and improving the planning efficiency and dynamic response capability for future tasks.

[0161] C3. Experimental Comparison and Result Analysis

[0162] C31. Experimental Conditions

[0163] In a dynamic warehousing scenario, the performance of the traditional A* algorithm and the combined A* and SAC algorithms provided by this invention are compared. The experimental scenario consists of 50 simulations, with dynamic obstacles moving randomly at different speeds and the task target point and starting point being randomly distributed.

[0164] C32, Experimental Indicators

[0165] Task completion time: The average time from task assignment to robot reaching the target point.

[0166] Obstacle avoidance success rate: The percentage of times a robot successfully avoids obstacles in a dynamic environment.

[0167] Path length: The average length of the path the robot takes to complete a task.

[0168] System stability: The consistency and reliability of the robot's path planning across multiple tasks.

[0169] C33, Experimental Results

[0170] Task completion time: The task completion time of the technical solution provided by this invention is significantly lower than that of the traditional A* algorithm, with an average reduction of about 30%, indicating that the path planning efficiency is higher.

[0171] Obstacle avoidance success rate: The obstacle avoidance success rate of this invention reaches over 95%, which is significantly better than the obstacle avoidance performance of the traditional A* algorithm in dynamic scenes.

[0172] Path length: The path length generated by this invention is shorter than that of the traditional A* algorithm, and the path is smoother, reducing unnecessary travel turns.

[0173] System stability: The success rate of this invention reached 100% in 50 experiments, while the traditional A* algorithm caused some tasks to fail due to the static nature of the path.

[0174] C34. Technical Effects and Application Value

[0175] Experimental verification shows that this invention combines the global path planning capability of the A* algorithm with the local optimization and dynamic adaptation capability of the SAC algorithm, which not only significantly improves path planning efficiency but also solves the problem of insufficient adaptability of traditional methods in dynamic environments. Its technical performance is superior to using the A* algorithm alone, and it has broad application value, especially suitable for scenarios such as dynamic warehousing, intelligent logistics, and autonomous driving.

[0176] In summary, the present invention has the following advantages:

[0177] (1) Improve the real-time performance and adaptability of path planning: By combining the global planning capability of the A* algorithm and the local optimization capability of the SAC algorithm, the problem of slow response and limitations of traditional methods in dynamic environments is solved, thus meeting the requirements of real-time performance and adaptability.

[0178] (2) Improve planning capabilities in high-dimensional environments: Through the reinforcement learning mechanism of the SAC algorithm, it can effectively process high-dimensional state spaces, dynamically perceive multi-dimensional changes in complex environments, accurately plan paths, and improve the robot's navigation capabilities and adaptability in high-dimensional environments.

[0179] (3) Global path optimization planning capability: The A* algorithm and the SAC algorithm are deeply integrated to ensure the global rationality of the path while having the ability to dynamically adjust, which makes up for the limitations of a single algorithm and makes path planning more intelligent and efficient.

[0180] (4) Reduce development and maintenance costs: The modular design, combined with standardized interfaces, enables efficient interaction, reduces technical complexity and maintenance costs, facilitates system expansion and upgrades, and has broad application prospects and promotional value.

[0181] The above specific embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for intelligent navigation planning of a mobile robot, characterized in that, The method comprises the following steps: collecting environmental information in real time, wherein the environmental information comprises static map information and real-time environmental information; planning a global path by using a global path planning algorithm according to the static map information; the global path optimization and adjustment adopts a soft actor critic algorithm, and the process comprises the following steps: inputting the global path and real-time environmental information contained in the environmental information; defining a state space S and an action space A, and determining information contained in the state space S and the action space A according to the environmental information; determining a dynamic reward function based on the information contained in the state space S and the action space A; the dynamic reward function is used to update the relationship between the robot and dynamic obstacles; wherein the dynamic reward function comprises a reward for approaching a target point, a punishment for approaching a dynamic obstacle, and a punishment for deviating from the global path; the path planning strategy is updated based on the dynamic reward function and the soft actor critic algorithm; and the global path is locally optimized based on the updated path planning strategy to obtain an optimized path; the dynamic reward function is determined according to the environmental information, and a specific dynamic reward function is determined by calculation based on the following formula: wherein, is a distance of the robot to a target point, is a distance of the robot to a nearest obstacle, is an offset distance of the robot to a global path, is a weight parameter; the weight parameter is set respectively according to path planning requirements for adjustment; the update of the path planning strategy comprises optimization of a strategy network and optimization of a value function; Policy network optimization: maximize the objective function, whose expression is: ; wherein, is a desired function, is an environment state at a current time, is an action selected according to a policy at state is a dynamic reward function; value function optimization: minimizing Bellman residual to update parameters of the strategy network; in combination with the real-time environmental information, the global path is optimized and adjusted by using a local path optimization algorithm, the global path is adjusted in real time to avoid dynamic obstacles, and an optimized path is obtained; generating a control instruction according to the optimized path, and driving the robot to move according to the optimized path through the control instruction.

2. The intelligent navigation planning method of claim 1, wherein, The action space A comprises a movement direction and a speed that can be selected by the robot; and the state space S comprises a current position, a speed and local environmental features of the robot.

3. The intelligent navigation planning method of claim 1, wherein, The real-time environmental information comprises road condition information, dynamic obstacle information and a movement state of the robot. The static map information comprises a current position, a target position and a global environmental map of the robot; and the control instruction comprises a navigation instruction, an avoidance priority instruction and an energy-saving mode instruction.

4. The intelligent navigation planning method of claim 1, wherein, The global path is planned by using an A-star algorithm.

5. The intelligent navigation planning method of claim 1, wherein, The intelligent navigation planning method further comprises a navigation experience optimization strategy, and the navigation experience optimization strategy comprises the following steps: after completing each task, the robot stores state, action and reward information in a shared experience pool in a path planning strategy process, and gradually optimizes a strategy network of the soft actor critic algorithm.

6. An intelligent navigation planning system for a mobile robot, characterized by The method comprises the following steps: an environmental perception module for collecting environmental information in real time, wherein the environmental information comprises static map information and real-time environmental information; a global path planning module for receiving the static map information sent by the environmental perception module, and generating a global path by using a global path planning algorithm; a local path optimization module for receiving real-time environmental information sent by the environmental perception module, optimizing and adjusting the global path by using a local path optimization algorithm, and adjusting the global path in real time to avoid dynamic obstacles to obtain an optimized path. The robot control module is configured to receive the optimized path sent by the local path optimization module, generate a control instruction according to the optimized path, and drive the robot to move according to the optimized path through the control instruction.

7. The intelligent navigation planning system of claim 6, wherein, The environment perception module comprises a laser radar and a depth camera, and is configured to collect static obstacle information, dynamic obstacle information and robot state information around the robot.

Images