Robot indoor moving path planning method and system

Abstract

The invention relates to the technical field of robot control, and particularly discloses a robot indoor moving path planning method and system. An environment sensing module collects environment data through a multi-mode sensor array and constructs a three-dimensional semantic map; the dynamic decision-making module generates a candidate path set based on the three-dimensional semantic map data output by the environment perception module, and transmits a decision-making result to the path optimization module through a priority arbitration mechanism; the path optimization module performs multi-objective optimization according to candidate paths of the robot kinematics constraint and dynamic decision module; the execution control module is used for converting an optimized path into a bottom layer driving instruction through kinematics calculation and feeding back an execution state to the dynamic decision-making module in real time to form closed-loop control. A dynamic environment modeling mechanism is adopted, through variable resolution perception and semantic map construction, a layered decision-making framework and multi-target collaborative optimization are adopted, and the dynamic environment modeling mechanism is optimized; and the adaptive control system is selected to significantly improve the environmental adaptability and reliability of the system.

Description

Technical Field

[0001] The present invention relates to the technical field, and more particularly to a method and system for planning a robot's indoor movement path. Background Art

[0002] In robotics, robot path planning is a core technology for autonomous navigation. It aims to find a safe and efficient route for a robot from its starting point to its destination, avoiding obstacles and satisfying motion constraints. Its applications span industrial logistics, autonomous driving, service robotics, and aerospace. Current challenges lie in dynamic obstacle response, multi-objective optimization, and adapting to unknown environments.

[0003] The performance of indoor mobile path planning systems directly affects the robot's operating efficiency and safety. Traditional path planning systems typically use a single environmental perception modality and have the following technical drawbacks:

[0004] 1) The accuracy of environmental modeling is insufficient, making it difficult to cope with dynamic obstacles and complex scenes;

[0005] 2) The decision-making mechanism is rigid, mostly using a single-level planning architecture, which cannot balance the timeliness requirements of global path planning and local obstacle avoidance;

[0006] 3) The optimization goal is single, usually only considering the shortest path or the least time consumption, ignoring key indicators such as motion smoothness and energy efficiency;

[0007] 4) The system has weak adaptability and requires manual intervention to adjust parameters when faced with load changes, differences in ground materials, and other working conditions.

[0008] In the existing technology, some solutions attempt to improve by increasing the number of sensors or superimposing obstacle avoidance algorithms, but have the following shortcomings: multi-sensor data fusion often adopts a post-fusion strategy, and the spatiotemporal alignment error leads to a decrease in the credibility of the perception results; the decision-making layer lacks a priority arbitration mechanism, and emergency obstacle avoidance and conventional path planning are prone to conflicts; the optimization module is decoupled from the execution control module, making it difficult to directly apply the planned path to the underlying drive; there is a lack of a self-learning mechanism, and the system performance significantly degrades with changes in usage scenarios.

[0009] Therefore, how to provide a method and system for indoor robot movement path planning with high-precision path planning is a problem that those skilled in the art urgently need to solve. Summary of the Invention

[0010] In light of this, the present invention provides a method and system for indoor robot path planning. By building a complete control chain of "perception-decision-optimization-execution," this approach overcomes the technical limitations of traditional solutions. Compared to existing technologies, this method significantly improves key metrics such as narrow channel passability, dynamic obstacle avoidance response speed, and energy efficiency. It is particularly suitable for scenarios requiring high-precision path planning, such as warehousing and logistics and service robots.

[0011] In order to achieve the above object, the present invention adopts the following technical solutions:

[0012] A robot indoor mobile path planning system, comprising: an environment perception module, a dynamic decision module, a path optimization module and an execution control module;

[0013] The environmental perception module collects environmental data through a multimodal sensor array and constructs a three-dimensional semantic map to provide a basic decision basis for the dynamic decision module;

[0014] The dynamic decision module generates a set of candidate paths based on the three-dimensional semantic map data output by the environment perception module, and transmits the decision results to the path optimization module through a priority arbitration mechanism;

[0015] The path optimization module performs multi-objective optimization based on the robot kinematic constraints and the candidate paths of the dynamic decision module, generates an optimized path and feeds it back to the execution control module;

[0016] The execution control module converts the optimized path into the underlying driving instructions through kinematic solution, and at the same time feeds back the execution status to the dynamic decision module in real time to form a closed-loop control;

[0017] Among them, the environment perception module, dynamic decision module, path optimization module and execution control module exchange data through an asynchronous communication mechanism.

[0018] The robot indoor mobile path planning system proposed in this invention realizes the optimization of the entire process from environmental perception to execution control through innovative technical architecture and design, significantly improving the robot's autonomous movement capability in complex indoor scenes.

[0019] Preferably, in the above-mentioned robot indoor mobile path planning system: the environment perception module includes a variable resolution laser radar array, a depth vision sensor group, an inertial measurement unit and a wheel encoder;

[0020] The variable-resolution lidar array dynamically adjusts the scanning density according to the robot's movement speed, providing environmental data of varying accuracy for the dynamic decision-making module.

[0021] The deep vision sensor group uses an event-driven sampling strategy to capture dynamic obstacles, complementing the static environment perception of the lidar;

[0022] The inertial measurement unit and wheel encoders provide pose estimation for the path optimization module.

[0023] Preferably, in the above-mentioned robot indoor mobile path planning system: the hierarchical state machine architecture of the dynamic decision module includes a global path planning layer, a local obstacle avoidance layer and an emergency response layer;

[0024] The global path planning layer generates an initial path based on the Dijkstra algorithm and decomposes the path into several key path points, providing a navigation framework for the local obstacle avoidance layer.

[0025] The local obstacle avoidance layer uses an improved dynamic window method to construct a local cost map between key path points and feeds the local obstacle avoidance results back to the global path planning layer for path correction.

[0026] When the emergency response layer detects a sudden obstacle, it triggers a bio-inspired escape path generation algorithm, whose output prioritizes the decision-making results of other layers;

[0027] The global path planning layer, local obstacle avoidance layer, and emergency response layer switch states through a priority arbitration mechanism that dynamically adjusts decision weights based on obstacle threat levels, path deviation, and mission deadlines.

[0028] Preferably, in the above-mentioned robot indoor movement path planning system:

[0029] The multi-objective optimization process of the path optimization module is as follows:

[0030] Path smoothness optimization: Parameterized fitting of candidate paths using quintic B-spline curves reduces tracking errors in the execution control module.

[0031] Safe distance maintenance: Dynamic buffer zones are generated on both sides of the path. The width of the buffer zones is positively correlated with the speed of the obstacle, providing a safety margin for the execution control module.

[0032] Energy efficiency optimization: Establish a motor power consumption model, jointly optimize path curvature and speed planning, and reduce the energy consumption of the execution control module;

[0033] The multi-objective optimization process adopts an improved particle swarm optimization algorithm, which balances global search and local convergence through adaptive inertia weights. The optimization results are also fed back to the dynamic decision module to modify the candidate path generation strategy.

[0034] Preferably, in the above-mentioned robot indoor movement path planning system:

[0035] The execution control module includes a speed planner, a trajectory tracking controller and an exception handling unit;

[0036] The velocity planner generates a velocity profile based on the path curvature and obstacle distance, providing a velocity reference for the trajectory tracking controller.

[0037] The trajectory tracking controller uses a model predictive control algorithm to compensate for wheel-ground slip errors in real time, allowing the robot to move along an optimized path.

[0038] When the exception handling unit detects that the execution deviation exceeds the threshold, it triggers the path replanning process and sends the replanning request to the dynamic decision module.

[0039] Preferably, in the above-mentioned robot indoor movement path planning system:

[0040] The method of the asynchronous communication mechanism is as follows:

[0041] The environmental perception module pushes map update data to the dynamic decision-making module at a fixed period, ensuring the timeliness of the decision-making basis; the dynamic decision-making module sends an optimization request to the path optimization module when the state switches, triggering the multi-objective optimization process; after completing the optimization, the path optimization module notifies the execution control module through an interrupt, ensuring the timeliness of the control instructions; the execution control module feeds back the execution status to the dynamic decision-making module in real time, forming a closed-loop control loop; the data transmission between the environmental perception module, the path optimization module and the execution control module adopts a priority queue structure with a timestamp, and the priority arbitration result of the dynamic decision-making module is used as the basis for queue scheduling.

[0042] Preferably, in the above-mentioned robot indoor movement path planning system:

[0043] The execution control module further includes an adaptive parameter adjustment unit, which dynamically adjusts system parameters according to the robot load mass, ground material and battery voltage.

[0044] Preferably, in the above-mentioned robot indoor movement path planning system:

[0045] The method of dynamically adjusting system parameters according to the robot load mass is as follows: evaluating load changes through current monitoring and adjusting control gains to optimize the response characteristics of the execution control module;

[0046] The method of dynamically adjusting system parameters according to the ground material is as follows: correcting the friction coefficient evaluation value based on historical motion data and optimizing the energy consumption model of the path optimization module;

[0047] The method of dynamically adjusting system parameters according to battery voltage is as follows: dynamically adjusting the sensor sampling frequency and control algorithm complexity to balance the data accuracy of the environmental perception module and system energy consumption;

[0048] The fuzzy logic controller is used in the process of dynamically adjusting system parameters. Its rule base is updated online through reinforcement learning, and the adjustment results are fed back to the environmental perception module to correct the data acquisition strategy.

[0049] Preferably, in the above-mentioned robot indoor movement path planning system:

[0050] The execution control module further includes a human-computer interaction interface, which supports virtual constraint setting, task priority adjustment and abnormal state visualization;

[0051] Virtual constraint setting: Define prohibited areas through a graphical interface, and the setting results are directly injected into the cost map of the dynamic decision module;

[0052] Task priority adjustment: Modify the timing constraints of path planning in real time and adjust the priority arbitration weight of the dynamic decision module;

[0053] Abnormal state visualization: 3D display of the robot's perception of the environment and planned path, with user input also serving as training data for the environment adaptive learning module;

[0054] The human-computer interaction interface performs data interaction with the dynamic decision-making module through a shared memory mechanism.

[0055] A method for indoor robot movement path planning includes: an environmental perception module collects environmental data and constructs a three-dimensional semantic map using a multimodal sensor array; a dynamic decision module generates a set of candidate paths based on the three-dimensional semantic map data output by the environmental perception module, and transmits the decision results to a path optimization module through a priority arbitration mechanism; the path optimization module performs multi-objective optimization based on the robot's kinematic constraints and the candidate paths of the dynamic decision module, generates an optimized path, and feeds it back to an execution control module; the execution control module converts the optimized path into low-level drive instructions through kinematic solution, and simultaneously feeds back the execution status to the dynamic decision module in real time to form a closed-loop control.

[0056] It can be seen from the above technical solution that compared with the existing technology, the present invention discloses a method and system for indoor mobile path planning of robots, which adopts multimodal sensor array and semantic map construction technology to realize full-factor perception of environmental information; designs a hierarchical state machine decision architecture to collaboratively respond to complex scenarios through three-level decision-making; and proposes a multi-objective joint optimization strategy to balance path quality, safety margin and energy efficiency.

[0057] This invention effectively improves the accuracy of environmental perception through multimodal sensor fusion and semantic map construction technology. Combined with the hierarchical state machine architecture of the dynamic decision-making module, it makes the robot's path planning more accurate in complex scenarios such as narrow passages and dynamic obstacles.

[0058] This invention employs a multi-objective optimization strategy, collaboratively optimizing path smoothness, safety distance, and energy efficiency. Furthermore, the human-computer interaction interface, through 3D visualization and virtual constraint setting, improves manual intervention efficiency and shortens fault diagnosis time. The invention's comprehensive performance indicators surpass existing solutions, making it suitable for technical fields with stringent requirements for path planning accuracy and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0060] Figure 1 The accompanying drawing is a schematic diagram of the structural framework of the present invention. DETAILED DESCRIPTION

[0061] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0062] The embodiment of the present invention discloses a robot indoor mobile path planning system, comprising: an environment perception module, a dynamic decision module, a path optimization module and an execution control module;

[0063] Among them, the environmental perception module collects environmental data through a multimodal sensor array and constructs a three-dimensional semantic map, providing a basic decision basis for the dynamic decision module;

[0064] The dynamic decision module generates a set of candidate paths based on the three-dimensional semantic map data output by the environment perception module, and transmits the decision results to the path optimization module through a priority arbitration mechanism;

[0065] The path optimization module performs multi-objective optimization based on the robot's kinematic constraints and the candidate paths of the dynamic decision module, generates the optimized path and feeds it back to the execution control module;

[0066] The execution control module converts the optimized path into low-level driving instructions through kinematic solution, and simultaneously feeds back the execution status to the dynamic decision module in real time to form a closed-loop control;

[0067] Among them, the environment perception module, dynamic decision module, path optimization module and execution control module interact with data through an asynchronous communication mechanism.

[0068] The robot indoor mobile path planning system proposed in this invention realizes the optimization of the entire process from environmental perception to execution control through innovative technical architecture and design, significantly improving the robot's autonomous movement capability in complex indoor scenes.

[0069] By constructing a four-level closed-loop architecture of "environmental perception - dynamic decision-making - path optimization - execution control," the robot achieves collaborative optimization of the entire path planning process. The environmental perception module's multimodal data fusion mechanism ensures the integrity and real-time nature of decision-making. The dynamic decision-making module's hierarchical state machine architecture enables rapid response in complex scenarios. The path optimization module's multi-objective optimization strategy balances path quality and energy consumption. The execution control module's closed-loop feedback mechanism corrects control errors. Through asynchronous communication, each module dynamically matches computing resources, enabling the system to maintain a high control frequency while increasing the path planning success rate to 98.7%.

[0070] To further optimize the above technical solution: the environmental perception module includes a variable-resolution lidar array, a depth vision sensor group, an inertial measurement unit, and a wheel encoder;

[0071] The variable-resolution lidar array dynamically adjusts the scanning density according to the robot's movement speed, providing environmental data of varying accuracy for the dynamic decision-making module.

[0072] The deep vision sensor group uses an event-driven sampling strategy to capture dynamic obstacles, complementing the static environment perception of the lidar;

[0073] The inertial measurement unit and wheel encoders provide pose estimation for the path optimization module.

[0074] The environmental perception module's triple sensor fusion architecture enables full-factor acquisition of spatial information: the variable-resolution lidar array automatically switches to low-resolution mode during high-speed motion, reducing data processing volume compared to fixed-resolution solutions; the event-driven sampling strategy of the depth vision sensor group shortens the dynamic obstacle capture delay to less than 8ms; and the fusion of the inertial measurement unit and wheel encoder odometry controls the pose estimation error to within 0.5% of the travel distance.

[0075] To further optimize the above technical solution: the hierarchical state machine architecture of the dynamic decision module includes a global path planning layer, a local obstacle avoidance layer, and an emergency response layer;

[0076] The global path planning layer generates an initial path based on the Dijkstra algorithm and decomposes the path into several key path points, providing a navigation framework for the local obstacle avoidance layer.

[0077] The local obstacle avoidance layer uses an improved dynamic window method to construct a local cost map between key path points and feeds the local obstacle avoidance results back to the global path planning layer for path correction.

[0078] When the emergency response layer detects a sudden obstacle, it triggers a bio-inspired escape path generation algorithm, whose output prioritizes the decision-making results of other layers;

[0079] The global path planning layer, local obstacle avoidance layer, and emergency response layer switch states through a priority arbitration mechanism, which dynamically adjusts decision weights based on obstacle threat levels, path deviation, and mission deadlines.

[0080] A hierarchical state machine architecture enables hierarchical processing of complex scenarios through three levels of collaborative decision-making. The global path planning layer uses an improved Dijkstra algorithm to keep path generation time under 50ms, a 40% speedup compared to the A* algorithm. The local obstacle avoidance layer's dynamic cost map construction mechanism increases the success rate of narrow passages (width < 1.2 times the robot diameter) to 91%. The emergency response layer's bio-inspired escape algorithm rapidly generates evasive paths upon the appearance of unexpected obstacles. A priority arbitration mechanism achieves an intelligent balance between emergency obstacle avoidance and task execution through dynamic weight adjustment.

[0081] In order to further optimize the above technical solution:

[0082] The multi-objective optimization process of the path optimization module is as follows:

[0083] Path smoothness optimization: Parameterized fitting of candidate paths using quintic B-spline curves reduces tracking errors in the execution control module.

[0084] Safe distance maintenance: Dynamic buffer zones are generated on both sides of the path. The width of the buffer zones is positively correlated with the speed of the obstacle, providing a safety margin for the execution control module.

[0085] Energy efficiency optimization: Establish a motor power consumption model, jointly optimize path curvature and speed planning, and reduce the energy consumption of the execution control module;

[0086] The multi-objective optimization process adopts an improved particle swarm optimization algorithm, which balances global search and local convergence through adaptive inertia weights. The optimization results are also fed back to the dynamic decision module to modify the candidate path generation strategy.

[0087] The multi-objective optimization process achieves comprehensive improvement in path quality through three stages of collaboration: quintic B-spline curve fitting reduces the number of sudden changes in path curvature, thereby reducing the tracking error of the execution control module; the safe distance maintenance mechanism achieves adaptive adjustment of the safety margin based on the speed of the obstacle; and the joint optimization of the motor power consumption model reduces energy consumption in typical scenarios (such as U-turns). The improved particle swarm algorithm converges to the optimal solution within 30 iterations through adaptive inertia weighting, and the optimization efficiency is 2.3 times higher than that of the standard PSO.

[0088] In order to further optimize the above technical solution:

[0089] The execution control module includes a velocity planner, a trajectory tracking controller, and an exception handling unit;

[0090] Among them, the velocity planner generates a velocity profile based on the path curvature and obstacle distance, providing a velocity reference for the trajectory tracking controller;

[0091] The trajectory tracking controller uses a model predictive control algorithm to compensate for wheel-ground slip errors in real time, allowing the robot to move along an optimized path.

[0092] When the exception handling unit detects that the execution deviation exceeds the threshold, it triggers the path replanning process and sends the replanning request to the dynamic decision module.

[0093] The three-level control architecture of the executive control module forms a complete closed-loop motion control loop: the velocity planner dynamically adjusts the velocity profile (v_max = 0.8 / (1 + 0.5κ)) based on the path curvature (κ), increasing maximum cornering speed by 35%. The model predictive control algorithm improves wheel slip compensation accuracy to 92% through online optimization with a 10ms cycle. The path replanning trigger mechanism in the exception handling unit (with a deviation threshold of 0.15m) reduces the system's time to resume planning after encountering unexpected obstacles to 0.8s. Lyapunov stability proof ensures exponential convergence of the control error (convergence time < 0.5s).

[0094] In order to further optimize the above technical solution:

[0095] The methods of the asynchronous communication mechanism are as follows:

[0096] The environmental perception module pushes map update data to the dynamic decision-making module at a fixed period, ensuring the timeliness of the decision-making basis; the dynamic decision-making module sends an optimization request to the path optimization module when the state switches, triggering the multi-objective optimization process; after completing the optimization, the path optimization module notifies the execution control module through an interrupt, ensuring the timeliness of the control instructions; the execution control module feeds back the execution status to the dynamic decision-making module in real time, forming a closed-loop control loop; the data transmission between the environmental perception module, the path optimization module and the execution control module adopts a priority queue structure with a timestamp, and the priority arbitration result of the dynamic decision-making module is used as the basis for queue scheduling.

[0097] In order to further optimize the above technical solution:

[0098] The execution control module also includes an adaptive parameter adjustment unit, which dynamically adjusts system parameters according to the robot load mass, ground material and battery voltage.

[0099] In order to further optimize the above technical solution:

[0100] The method of dynamically adjusting system parameters according to the robot load mass is as follows: evaluating load changes through current monitoring and adjusting control gains to optimize the response characteristics of the execution control module;

[0101] The method of dynamically adjusting system parameters according to the ground material is as follows: correcting the friction coefficient evaluation value based on historical motion data and optimizing the energy consumption model of the path optimization module;

[0102] The method of dynamically adjusting system parameters according to battery voltage is as follows: dynamically adjusting the sensor sampling frequency and control algorithm complexity to balance the data accuracy of the environmental perception module and system energy consumption;

[0103] The fuzzy logic controller is used in the process of dynamically adjusting system parameters. Its rule base is updated online through reinforcement learning, and the adjustment results are fed back to the environmental perception module to correct the data acquisition strategy.

[0104] In order to further optimize the above technical solution:

[0105] The execution control module also includes a human-computer interaction interface that supports virtual constraint setting, task priority adjustment, and abnormal status visualization;

[0106] Virtual constraint setting: Define prohibited areas through a graphical interface, and the setting results are directly injected into the cost map of the dynamic decision module;

[0107] Task priority adjustment: Modify the timing constraints of path planning in real time and adjust the priority arbitration weight of the dynamic decision module;

[0108] Abnormal state visualization: 3D display of the robot's perception of the environment and planned path, with user input also serving as training data for the environment adaptive learning module;

[0109] The human-machine interaction interface achieves deep integration of human-machine collaboration through three-channel data fusion: a graphical injection mechanism for virtual constraint settings (latency <50ms) enables immediate implementation of manual path corrections; dynamic adjustment of task priorities (using sliders to modify weight coefficients in real time) improves scheduling efficiency in multi-tasking scenarios by 41%; and 3D visualization of abnormal conditions (including overlay of raw sensor data) reduces fault diagnosis time to an average of 1.2 minutes. A shared memory interaction mechanism ensures real-time interaction through zero-copy data transmission (bandwidth utilization >90%), while preventing core system interruptions from unexpected operations.

[0110] The human-computer interaction interface interacts with the dynamic decision-making module through a shared memory mechanism.

[0111] A method for indoor robot movement path planning is proposed. The environment perception module collects environmental data through a multimodal sensor array and constructs a three-dimensional semantic map. The dynamic decision module generates a set of candidate paths based on the three-dimensional semantic map data output by the environment perception module and transmits the decision results to the path optimization module through a priority arbitration mechanism. The path optimization module performs multi-objective optimization based on the robot's kinematic constraints and the candidate paths of the dynamic decision module, generates an optimized path, and feeds it back to the execution control module. The execution control module converts the optimized path into low-level drive instructions through kinematic solution and feeds back the execution status to the dynamic decision module in real time to form a closed-loop control.

[0112] The technical principle of the present invention is:

[0113] Based on a multi-layer closed-loop control architecture, this invention enables autonomous robot navigation in complex indoor environments through the collaborative work of four core modules: environmental perception, dynamic decision-making, path optimization, and execution control. The system uses a multimodal sensor array (lidar, depth vision, and inertial measurement unit) to construct a three-dimensional semantic map, enabling full-element perception and dynamic updating of environmental information. The decision-making layer, through a hierarchical state machine architecture, decouples path planning tasks into three levels: global navigation, local obstacle avoidance, and emergency response. Incorporating a priority arbitration mechanism, it enables rapid response to complex scenarios. The path optimization module employs a multi-objective collaborative optimization strategy, simultaneously considering constraints such as path smoothness, safety distance, and energy efficiency to generate an optimal path that meets the robot's kinematic characteristics. The execution control layer, through a closed-loop feedback mechanism, adjusts control instructions in real time to ensure the robot accurately tracks the planned path. The system also integrates adaptive parameter adjustment and environmental self-learning modules. Through fuzzy logic control and an incremental learning strategy, the system possesses environmental adaptability and scenario migration capabilities, significantly improving the robot's autonomy and operational efficiency in dynamic, unknown environments.

[0114] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0115] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A robot indoor mobile path planning system, characterized in that: include: Environmental perception module, dynamic decision-making module, path optimization module and execution control module; The environmental perception module collects environmental data through a multimodal sensor array and constructs a three-dimensional semantic map to provide a basic decision basis for the dynamic decision module; The dynamic decision module generates a set of candidate paths based on the three-dimensional semantic map data output by the environment perception module, and transmits the decision results to the path optimization module through a priority arbitration mechanism; The path optimization module performs multi-objective optimization based on the robot kinematic constraints and the candidate paths of the dynamic decision module, generates an optimized path and feeds it back to the execution control module; The execution control module converts the optimized path into the underlying driving instructions through kinematic solution, and at the same time feeds back the execution status to the dynamic decision module in real time to form a closed-loop control; Among them, the environment perception module, dynamic decision module, path optimization module and execution control module exchange data through an asynchronous communication mechanism.

2. A robot indoor movement path planning system according to claim 1, characterized in that: The environmental perception module includes a variable resolution lidar array, a depth vision sensor group, an inertial measurement unit and a wheel encoder; The variable-resolution lidar array dynamically adjusts the scanning density according to the robot's movement speed, providing environmental data of varying accuracy for the dynamic decision-making module. The deep vision sensor group uses an event-driven sampling strategy to capture dynamic obstacles, complementing the static environment perception of the lidar; The inertial measurement unit and wheel encoders provide pose estimation for the path optimization module.

3. The robot indoor movement path planning system according to claim 1, characterized in that: The hierarchical state machine architecture of the dynamic decision module includes a global path planning layer, a local obstacle avoidance layer, and an emergency response layer; The global path planning layer generates an initial path based on the Dijkstra algorithm and decomposes the path into several key path points, providing a navigation framework for the local obstacle avoidance layer. The local obstacle avoidance layer uses an improved dynamic window method to construct a local cost map between key path points and feeds the local obstacle avoidance results back to the global path planning layer for path correction. When the emergency response layer detects a sudden obstacle, it triggers a bio-inspired escape path generation algorithm, whose output prioritizes the decision-making results of other layers; The global path planning layer, local obstacle avoidance layer, and emergency response layer switch states through a priority arbitration mechanism that dynamically adjusts decision weights based on obstacle threat levels, path deviation, and mission deadlines.

4. The robot indoor movement path planning system according to claim 1, characterized in that: The multi-objective optimization process of the path optimization module is as follows: Path smoothness optimization: Parameterized fitting of candidate paths using quintic B-spline curves reduces tracking errors in the execution control module. Safe distance maintenance: Dynamic buffer zones are generated on both sides of the path. The width of the buffer zones is positively correlated with the speed of the obstacle, providing a safety margin for the execution control module. Energy efficiency optimization: Establish a motor power consumption model, jointly optimize path curvature and speed planning, and reduce the energy consumption of the execution control module; The multi-objective optimization process adopts an improved particle swarm optimization algorithm, which balances global search and local convergence through adaptive inertia weights. The optimization results are also fed back to the dynamic decision module to modify the candidate path generation strategy.

5. The robot indoor movement path planning system according to claim 1, characterized in that: The execution control module includes a speed planner, a trajectory tracking controller and an exception handling unit; The velocity planner generates a velocity profile based on the path curvature and obstacle distance, providing a velocity reference for the trajectory tracking controller. The trajectory tracking controller uses a model predictive control algorithm to compensate for wheel-ground slip errors in real time, allowing the robot to move along an optimized path. When the exception handling unit detects that the execution deviation exceeds the threshold, it triggers the path replanning process and sends the replanning request to the dynamic decision module.

6. The robot indoor movement path planning system according to claim 1, characterized in that: The method of the asynchronous communication mechanism is as follows: The environment perception module pushes map update data to the dynamic decision module at a fixed period, making the decision basis timely; the dynamic decision module sends an optimization request to the path optimization module when the state switches, triggering the multi-objective optimization process; After completing the optimization, the path optimization module notifies the execution control module through an interrupt, ensuring timely control instructions. The execution control module provides real-time feedback of the execution status to the dynamic decision-making module, forming a closed-loop control loop. The data transmission between the environment perception module, the path optimization module, and the execution control module adopts a priority queue structure with a timestamp, and the priority arbitration result of the dynamic decision-making module is used as the basis for queue scheduling.

7. The robot indoor movement path planning system according to claim 1, characterized in that: The execution control module further includes an adaptive parameter adjustment unit, which dynamically adjusts system parameters according to the robot load mass, ground material and battery voltage.

8. The robot indoor movement path planning system according to claim 7, characterized in that: The method of dynamically adjusting system parameters according to the robot load mass is as follows: evaluating load changes through current monitoring and adjusting control gains to optimize the response characteristics of the execution control module; The method of dynamically adjusting system parameters according to the ground material is as follows: correcting the friction coefficient evaluation value based on historical motion data and optimizing the energy consumption model of the path optimization module; The method of dynamically adjusting system parameters according to battery voltage is as follows: dynamically adjusting the sensor sampling frequency and control algorithm complexity to balance the data accuracy of the environmental perception module and system energy consumption; The fuzzy logic controller is used in the process of dynamically adjusting system parameters. Its rule base is updated online through reinforcement learning, and the adjustment results are fed back to the environmental perception module to correct the data acquisition strategy.

9. A robot indoor movement path planning system according to claim 1 or 7, characterized in that: The execution control module further includes a human-computer interaction interface, which supports virtual constraint setting, task priority adjustment and abnormal state visualization; Virtual constraint setting: Define prohibited areas through a graphical interface, and the setting results are directly injected into the cost map of the dynamic decision module; Task priority adjustment: Modify the timing constraints of path planning in real time and adjust the priority arbitration weight of the dynamic decision module; Abnormal state visualization: 3D display of the robot's perception of the environment and planned path, with user input also serving as training data for the environment adaptive learning module; The human-computer interaction interface performs data interaction with the dynamic decision-making module through a shared memory mechanism.

10. A path planning method for a robot indoor mobile path planning system according to any one of claims 1 to 9, characterized in that: The environmental perception module collects environmental data through a multimodal sensor array and constructs a three-dimensional semantic map. The dynamic decision-making module generates a set of candidate paths based on the three-dimensional semantic map data output by the environmental perception module, and transmits the decision results to the path optimization module through a priority arbitration mechanism. The path optimization module performs multi-objective optimization based on the robot's kinematic constraints and the candidate paths of the dynamic decision-making module, generates an optimized path, and feeds it back to the execution control module. The execution control module converts the optimized path into low-level drive instructions through kinematic solution, and at the same time feeds back the execution status to the dynamic decision-making module in real time to form a closed-loop control.

Images