Autonomous navigation method and system for intelligent inspection quadruped robot

Abstract

The invention discloses an autonomous navigation method and system for an intelligent inspection quadruped robot. The system comprises a sensing module, a navigation module, a control module and a communication module, the sensing module is used for environment sensing, and is responsible for collecting environment information and transmitting data to the navigation module for path planning; the navigation module is responsible for path planning and automatic obstacle avoidance, and after an optimal path is planned according to the sensing data, an instruction is sent to the control module; the control module is used for executing the control instruction and controlling the robot to move according to the navigation instruction; and meanwhile, the communication module realizes interaction with an external system and is responsible for data transmission and communication between the robot and other systems or equipment, so that cooperative work among the modules is ensured, and an efficient and accurate inspection task is realized. According to the invention, through fusion of the improved DWA algorithm and the A * algorithm, in combination with the high-precision sensing module and the powerful control module, autonomous navigation and dynamic obstacle avoidance of the robot in a complex environment are realized, and the efficiency and safety of an inspection task are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent robots, and in particular to an autonomous navigation method and system for an intelligent inspection quadruped robot. Background Art

[0002] With the acceleration of globalization, the demand for infrastructure and public services is also growing. As a new generation of automated equipment, intelligent inspection robots play an increasingly important role in management. Existing inspection robots are mostly wheeled or tracked. These types of robots have poor motion performance and stability in complex terrains and cannot meet the changing inspection needs in the environment. Quadruped robots have better adaptability and stability in complex terrains due to their unique gait and movement mode, making them a new choice for intelligent inspection.

[0003] The intelligent inspection quadruped robot mainly relies on the A* algorithm for path planning. Although this algorithm can effectively plan paths in static environments, it lacks flexibility and adaptability when facing dynamic obstacles, which can easily lead to path planning failures or interruption of inspection tasks. At the same time, the traditional dynamic window method is prone to local optimality when dealing with such problems.

[0004] Therefore, there is an urgent need for a method and system that can effectively improve the autonomous navigation capability of a quadruped robot so as to better adapt to the complex inspection environment and improve inspection efficiency and safety. Summary of the invention

[0005] In order to overcome the above technical problems, the purpose of the present invention is to provide an autonomous navigation method and system for an intelligent inspection quadruped robot. By integrating the improved DWA algorithm with the A* algorithm, combined with a high-precision perception module and a powerful control module, the robot can achieve autonomous navigation and dynamic obstacle avoidance in complex environments, thereby improving the efficiency and safety of inspection tasks.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is:

[0007] An autonomous navigation method and system for an intelligent inspection quadruped robot, comprising a perception module, a navigation module, a control module, and a communication module;

[0008] The perception module is used for environmental perception, responsible for collecting environmental information and transmitting data to the navigation module for path planning;

[0009] The navigation module is responsible for path planning and automatic obstacle avoidance. After planning the optimal path based on the perception data, it sends instructions to the control module.

[0010] The control module is used to execute control instructions and control the movement of the robot according to navigation instructions; at the same time, the communication module realizes interaction with external systems and is responsible for data transmission and communication between the robot and other systems or devices, ensuring the coordination between various modules to achieve efficient and accurate inspection tasks. The perception module includes a laser radar and a depth camera, which are used to obtain surrounding environment information in real time and form an environmental map;

[0011] The laser radar is installed on the top of the quadruped robot to maximize its field of view and detection capability;

[0012] The depth camera is installed on the face of the robot to better capture the three-dimensional information of the environment in front. The perception module improves the accuracy and reliability of environmental perception through multi-sensor data fusion technology.

[0013] The navigation module uses the improved DWA algorithm and the A* algorithm to fuse the environment map provided by the perception module, uses the A* algorithm to generate a global path, extracts the optimal node and outputs the global optimal path, and then uses the dynamic window method to perform local path planning between adjacent nodes after obtaining the global optimal node sequence, and performs dynamic obstacle avoidance and path adjustment in a local range to achieve efficient autonomous navigation;

[0014] An exponential function with a distance attenuation factor is introduced into the obstacle distance evaluation function of the DWA algorithm, as shown below:

[0015] f(x)=e -δx

[0016] Among them, x represents the actual distance between the quadruped robot and the obstacle, and δ is the distance attenuation coefficient. This method can reduce the impact of long-distance obstacles on path planning during the driving of the quadruped robot and increase the priority of close-distance obstacles to more accurately reflect the actual situation. Finally, the improved obstacle distance evaluation function is as follows:

[0017]

[0018] Among them, (x, y) is the coordinate of the starting point of the quadruped robot, (x ob ,y ob ) is the coordinate of the target point of the quadruped robot, r is the safety radius. It can be seen from the formula that the size of δ can adjust the degree of attention to different obstacles in path planning. A larger δ((0.5≤δ≤1()) strengthens the avoidance of close obstacles, and a smaller δ(0.1≤δ<0.5) maintains a higher speed at a long distance.

[0019] The improved DWA algorithm introduces a path evaluation sub-function path in the DWA algorithm, and the smaller the weight is, the closer it is to the global path planning, and vice versa, the more it deviates from the global path planning;

[0020] Then introduce the angle factor θ ob , which is the angle between the quadruped robot and the obstacle, θ ob is always less than 180°, and then θ ob It is added to the denominator of the trajectory evaluation function to dynamically adjust the weight of the sub-evaluation function, making the path planning more flexible in different situations.

[0021] The basic parameters of the improved DWA algorithm are set as follows: the maximum speed limit is set to 1m / s; the maximum angular velocity limit is set to 20° / s; the speed resolution is set to 0.01m / s; the angular velocity resolution is set to 1° / s; the acceleration limit is 0.2m / s 2 ; Angular acceleration is limited to 50° / s 2 The parameters of the evaluation function are: α=0.1, β=0.05, γ=0.2, ζ=0.5, and the prediction time period is set to 3.0s.

[0022] The improved trajectory evaluation function is as follows:

[0023]

[0024] Where: αPHead(v,ω) is the angle difference between the end point of the trajectory and the current target point; dist(v,ω) is the distance between the trajectory and the nearest obstacle; vel(v,ω) is the current speed evaluation function; path(v,ω) is the path evaluation subfunction, σ is the smoothing function; α, β, γ are weighting coefficients, θ ob is the angle factor.

[0025] The control module includes a motion control unit and a task management unit. The motion control unit is responsible for the coordination and control of the robot's quadruped motion. The task management unit issues execution instructions based on the path and obstacle avoidance information provided by the navigation module to achieve autonomous inspection tasks.

[0026] The communication module adopts wireless communication technology WiFi, 4G or 5G network to realize real-time data interaction between the quadruped robot and the cloud platform and inspection system, and supports remote monitoring and control.

[0027] A method for using an autonomous navigation system of an intelligent inspection quadruped robot comprises the following steps:

[0028] Step 1: Initialize the system:

[0029] Turn on the quadruped robot and start the perception module, navigation module, control module and communication module;

[0030] Step 2: Environmental Perception:

[0031] The perception module acquires environmental data in real time through lidar and depth camera and generates an environmental map;

[0032] Step 3: Global path planning: The navigation module uses the A* algorithm to generate a global path based on the environment map, extract the optimal nodes and output the global optimal path to obtain the global optimal node sequence;

[0033] Step 4: Local path planning and obstacle avoidance: During the local path planning process, the navigation module uses the improved DWA algorithm to perform dynamic obstacle avoidance and path adjustment;

[0034] Step 5: Motion Control:

[0035] The control module issues control instructions and coordinates the movement of the quadruped robot based on the path and obstacle avoidance information provided by the navigation module;

[0036] Execute navigation instructions to realize autonomous movement of the robot;

[0037] Monitor and adjust the robot's posture and gait to ensure smooth movement.

[0038] Step 6: Data Interaction:

[0039] The communication module realizes real-time data interaction between the quadruped robot and the cloud platform and inspection system through WiFi, 4G or 5G network, and supports remote monitoring and control;

[0040] Step 7: Task Execution:

[0041] Perform inspection tasks according to the instructions of the task management unit; monitor the progress of tasks in real time to ensure smooth completion of tasks; record and report data collected during the inspection process.

[0042] The specific steps of step 2 include:

[0043] The laser radar on the top of the quadruped robot is used to obtain point cloud data of the surrounding environment, and the point cloud data is subjected to noise reduction and filtering.

[0044] The depth camera on the face of the quadruped robot is used to obtain the depth image of the surrounding environment, and the image processing algorithm is used to identify obstacles and terrain features in the real-time environment;

[0045] Project the LiDAR point cloud data onto a two-dimensional plane to generate a two-dimensional point cloud projection map. Then convert the depth image data of the depth camera into corresponding three-dimensional points and project them onto the same two-dimensional plane. Align the point cloud projection map and the depth map projection map, perform data fusion, and ensure the spatial consistency of the two sensor data.

[0046] Initialize an empty two-dimensional grid map according to the predetermined environment size and resolution. Each grid represents a small area in the environment, usually a square. White grids represent idle states and black grids represent obstacles. Then use the Bayesian update rule to fuse the lidar and depth camera data into the grid map. The final output two-dimensional grid map will contain obstacle information in the environment for robot navigation and path planning.

[0047] The specific steps of step 3 include:

[0048] Combined with the data from the environment perception module, a two-dimensional grid map of the current environment is generated;

[0049] Set the robot's starting position and target position according to the task requirements and mark them on the environment map;

[0050] The A* algorithm is used to start from the starting node, gradually evaluate possible path nodes, calculate f(n) in each step, and select the node with the smallest f(n) for expansion until the target node is reached;

[0051] The A* algorithm of the present invention uses the Euclidean distance to define the heuristic function, which is:

[0052]

[0053] In the formula, n represents the current node of the robot, h(n) represents the distance from the current node to the target node, and n x Indicates the x-axis coordinate of the current node, n y represents the y-axis coordinate of the current node; g represents the robot's target node, g x Indicates the x-axis coordinate of the target node, g y Indicates the y-axis coordinate of the target node.

[0054] The cost function formula of the A* algorithm is:

[0055] f(n)=g(n)+h(n)

[0056] Where f(n) represents the estimated total cost from the starting node to the target node via node n, g(n) represents the actual distance from the starting node to node n in the state space, and h(n) represents the estimated cost from node n to the target node;

[0057] Finally, the node sequence in the expansion process is recorded, and the node sequence of the optimal path is finally output.

[0058] The specific steps of step 4 include:

[0059] An exponential function with a distance attenuation factor is introduced into the obstacle distance evaluation function of the DWA algorithm, as shown below:

[0060] f(x)=e -δx

[0061] In the formula, x represents the actual distance between the quadruped robot and the obstacle, and δ is the distance attenuation coefficient. This method can reduce the impact of long-distance obstacles on path planning during the driving of the quadruped robot and increase the priority of close-distance obstacles to more accurately reflect the actual situation. Finally, the improved obstacle distance evaluation function is as follows:

[0062]

[0063] Among them, (x, y) is the coordinate of the starting point of the quadruped robot, (x ob ,y ob ) is the coordinate of the target point of the quadruped robot, r is the safety radius. It can be seen from the formula that the size of δ can adjust the degree of attention to different obstacles in path planning. A larger δ ((0.5≤δ≤1()) strengthens the avoidance of close-range obstacles, and a smaller δ (0.1≤δ<0.5) maintains a higher speed at a long distance;

[0064] The improved DWA algorithm introduces a path evaluation sub-function path in the DWA algorithm, and the smaller the weight, the closer it is to the global path planning, and vice versa, the more it deviates from the global path planning;

[0065] Then introduce the angle factor θ ob , which is the angle between the quadruped robot and the obstacle, θ ob is always less than 180°, and then θ ob It is added to the denominator of the trajectory evaluation function to dynamically adjust the weight of the sub-evaluation function, making the path planning more flexible in different situations;

[0066] The basic parameters of the improved DWA algorithm are set as follows: the maximum speed limit is set to 1m / s; the maximum angular velocity limit is set to 20° / s; the speed resolution is set to 0.01m / s; the angular velocity resolution is set to 1° / s; the acceleration limit is 0.2m / s 2 ; Angular acceleration is limited to 50° / s 2 The parameters of the evaluation function are: α=0.1, β=0.05, γ=0.2, ζ=0.5, and the prediction time period is set to 3.0s.

[0067] The improved trajectory evaluation function is as follows:

[0068]

[0069] Where: αPHead(v,ω) is the angle difference between the end point of the trajectory and the current target point; dist(v,ω) is the distance between the trajectory and the nearest obstacle; vel(v,ω) is the current speed evaluation function; path(v,ω) is the path evaluation subfunction, σ is the smoothing function; α, β, γ are weighting coefficients, θ ob is the angle factor.

[0070] It is best to refine the path between a series of key nodes extracted from the global optimal path based on the improved DWA algorithm, and dynamically adjust the path through the current speed, direction and obstacle information of the quadruped robot, so that the robot can flexibly avoid close obstacles and adjust the motion trajectory in real time to ensure that the quadruped robot moves efficiently and safely along the optimal path.

[0071] The step 5 specifically includes:

[0072] Control instructions are issued: The control module receives the path information and obstacle avoidance information provided by the navigation module and the perception module in real time; based on the received information, it calculates the necessary walking instructions, including the adjustment of speed, direction and gait; the control instructions are sent to the quadruped robot drive system through the internal communication protocol;

[0073] Execute navigation instructions: After receiving the control instructions, the drive system drives the joints and motors of the quadruped robot to move according to the specified motion mode; through feedback systems such as encoders and angle sensors, the accuracy and stability of the execution process are monitored to ensure the correct implementation of the instructions;

[0074] Posture and gait adjustment: Monitor the posture data of the quadruped robot in real time and collect it through gyroscope attitude sensors and accelerometers; dynamically adjust the gait and body posture of the quadruped robot based on the monitoring data, and use the MPC algorithm (model predictive control) to fine-tune the joint angle and step frequency to ensure that the quadruped robot can move smoothly on different terrains.

[0075] The step 7 specifically includes:

[0076] Inspection task execution: receiving inspection tasks assigned by the task management unit to the quadruped robot, including the task location, target inspection items and scheduled timetable; breaking down the overall task into several subtasks, inspecting specific equipment and collecting specific environmental data, and arranging the time sequence and priority;

[0077] Task monitoring and adjustment: During the task execution, the position, status and environmental changes of the quadruped robot are monitored in real time; according to the actual situation, if the quadruped robot deviates from the planned path or the task is delayed, the system automatically adjusts the travel path or rearranges the execution order of subtasks;

[0078] Data collection and reporting: When performing inspection tasks, the required data, such as environmental information images, are collected through the sensor system; the collected data is integrated and an inspection report is generated, which is sent to the central control system in real time for analysis and recording.

[0079] Through the above steps, the autonomous navigation and dynamic obstacle avoidance of the intelligent inspection quadruped robot in complex environments can be realized, thus improving the efficiency and safety of inspection tasks.

[0080] Beneficial effects of the present invention:

[0081] 1. High navigation accuracy: Through multi-sensor data fusion and improved DWA algorithm and A* algorithm fusion, the A* algorithm is used to plan the optimal path on the global environment map, extract a series of key nodes, and the improved DWA algorithm is refined between these nodes. By considering the robot's current speed, direction and obstacle information, the path is dynamically adjusted, so that the robot can flexibly avoid close obstacles. The robot can achieve autonomous navigation and dynamic obstacle avoidance in complex environments. The navigation accuracy, path planning efficiency and automatic obstacle avoidance capabilities of the quadruped robot are improved.

[0082] 2. Real-time monitoring: The communication module realizes real-time data interaction with the cloud platform and inspection system, supports remote monitoring and control, and improves management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is the system architecture diagram of the intelligent inspection quadruped robot of the present invention.

[0084] Figure 2 This is a schematic diagram of the structure of the perception module of the present invention.

[0085] Figure 3 This is a flow chart of the navigation module of the present invention.

[0086] Figure 4 This is an experimental diagram of the fusion path planning algorithm of the present invention.

[0087] Figure 5 This is the posture angle diagram of the improved DWA algorithm of the present invention.

[0088] Figure 6 This is a working diagram of the control module of the present invention.

[0089] Figure 7 This is a schematic diagram of data interaction of the communication module of the present invention. DETAILED DESCRIPTION

[0090] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0091] like Figure 1 As shown in the figure, the autonomous navigation system of the intelligent inspection quadruped robot includes a perception module, a navigation module, a control module and a communication module. The quadruped robot perceives the environment through the perception module, the navigation module is responsible for path planning and automatic obstacle avoidance of the quadruped robot, the control module is used to execute control instructions, and the communication module realizes the interaction between the quadruped robot and the external system.

[0092] like Figure 2 As shown in the figure, the perception module includes a laser radar and a depth camera, which are used to obtain surrounding environment information in real time and form an environmental map. The accuracy and reliability of environmental perception are improved through multi-sensor data fusion technology.

[0093] like Figure 3 As shown in the figure, the navigation module is based on the environment map provided by the perception module, and adopts the improved DWA algorithm to fuse with the A* algorithm. The A* algorithm is used to generate the global path. After obtaining the global optimal node sequence, a series of key nodes are extracted, and the improved DWA algorithm is used to refine between these nodes. By considering the current speed, direction and obstacle information of the robot, the path is dynamically adjusted to improve the robot's autonomous navigation ability in complex environments.

[0094] like Figure 4 As shown in the figure, the improved DWA algorithm combined with the A* algorithm can successfully avoid random obstacles and reach the destination. It can not only correct the local path according to the global path, but also meet the requirements of the quadruped robot to avoid collisions during the pathfinding process. It can be seen that the feasibility and effectiveness of the algorithm have certain advantages.

[0095] like Figure 5 As shown in the figure, the improved DWA algorithm has a smoother posture angle and fewer peaks during path planning, which can effectively improve the robot's mobility efficiency and automatic obstacle avoidance capabilities.

[0096] like Figure 6 As shown, the control module includes a motion control unit and a task management unit. The motion control unit is responsible for the coordination and control of the robot's quadruped motion. The task management unit issues execution instructions based on the path and obstacle avoidance information provided by the navigation module to achieve autonomous inspection tasks.

[0097] like Figure 7 As shown, the communication module uses wireless communication technology (such as WiFi, 4G / 5G network) to realize real-time data interaction between the robot and the cloud platform and inspection system, and supports remote monitoring and control of the quadruped robot.

[0098] A method for using an autonomous navigation system of an intelligent inspection quadruped robot comprises the following steps:

[0099] Step 1: Initialize the system:

[0100] Turn on the quadruped robot and start the perception module, navigation module, control module and communication module;

[0101] Step 2: Environmental Perception:

[0102] The perception module uses lidar and depth cameras to obtain real-time environmental data and generate an environmental map. The specific steps include:

[0103] The laser radar on the top of the quadruped robot is used to obtain point cloud data of the surrounding environment, and the point cloud data is subjected to noise reduction and filtering.

[0104] The depth camera on the face of the quadruped robot is used to obtain the depth image of the surrounding environment, and the image processing algorithm is used to identify obstacles and terrain features in the real-time environment;

[0105] Project the LiDAR point cloud data onto a two-dimensional plane to generate a two-dimensional point cloud projection map. Then convert the depth image data of the depth camera into corresponding three-dimensional points and project them onto the same two-dimensional plane. Align the point cloud projection map and the depth map projection map, perform data fusion, and ensure the spatial consistency of the two sensor data.

[0106] Initialize an empty two-dimensional grid map according to the predetermined environment size and resolution. Each grid represents a small area in the environment, usually a square. White grids represent idle states and black grids represent obstacles. Then use the Bayesian update rule to fuse the lidar and depth camera data into the grid map. The final output two-dimensional grid map will contain obstacle information in the environment for robot navigation and path planning.

[0107] Step 3: Global path planning: The navigation module uses the A* algorithm to generate a global path based on the environment map, extract the optimal nodes and output the global optimal path to obtain the global optimal node sequence;

[0108] The specific steps include:

[0109] Combined with the data from the environment perception module, a two-dimensional grid map of the current environment is generated;

[0110] Set the robot's starting position and target position according to the task requirements and mark them on the environment map;

[0111] The A* algorithm is used to start from the starting node, gradually evaluate possible path nodes, calculate f(n) in each step, and select the node with the smallest f(n) for expansion until the target node is reached;

[0112] The A* algorithm of the present invention uses the Euclidean distance to define the heuristic function, which is:

[0113]

[0114] In the formula, n represents the current node of the robot, h(n) represents the distance from the current node to the target node, and n x Indicates the x-axis coordinate of the current node, n y represents the y-axis coordinate of the current node; g represents the robot's target node, g x Indicates the x-axis coordinate of the target node, g y Indicates the y-axis coordinate of the target node.

[0115] The cost function formula of the A* algorithm is:

[0116] f(n)=g(n)+h(n)

[0117] Where f(n) represents the estimated total cost from the starting node to the target node via node n, g(n) represents the actual distance from the starting node to node n in the state space, and h(n) represents the estimated cost from node n to the target node;

[0118] Finally, the node sequence in the expansion process is recorded, and the node sequence of the optimal path is finally output.

[0119] Step 4: Local path planning and obstacle avoidance: During the local path planning process, the navigation module uses the improved DWA algorithm to perform dynamic obstacle avoidance and path adjustment;

[0120] The specific steps include:

[0121] An exponential function with a distance attenuation factor is introduced into the obstacle distance evaluation function of the DWA algorithm, as shown below:

[0122] f(x)=e -δx

[0123] In the formula, x represents the actual distance between the quadruped robot and the obstacle, and δ is the distance attenuation coefficient. This method can reduce the impact of long-distance obstacles on path planning during the driving of the quadruped robot and increase the priority of close-distance obstacles to more accurately reflect the actual situation. Finally, the improved obstacle distance evaluation function is as follows:

[0124]

[0125] Among them, (x, y) is the coordinate of the starting point of the quadruped robot, (x ob ,y ob ) is the coordinate of the target point of the quadruped robot, r is the safety radius. It can be seen from the formula that the size of δ can adjust the degree of attention to different obstacles in path planning. A larger δ ((0.5≤δ≤1()) strengthens the avoidance of close-range obstacles, and a smaller δ (0.1≤δ<0.5) maintains a higher speed at a long distance;

[0126] The improved DWA algorithm introduces a path evaluation sub-function path in the DWA algorithm, and the smaller the weight, the closer it is to the global path planning, and vice versa, the more it deviates from the global path planning;

[0127] Then introduce the angle factor θ ob , which is the angle between the quadruped robot and the obstacle, θ ob is always less than 180°, and then θ ob It is added to the denominator of the trajectory evaluation function to dynamically adjust the weight of the sub-evaluation function, making the path planning more flexible in different situations;

[0128] The basic parameters of the improved DWA algorithm are set as follows: the maximum speed limit is set to 1m / s; the maximum angular velocity limit is set to 20° / s; the speed resolution is set to 0.01m / s; the angular velocity resolution is set to 1° / s; the acceleration limit is 0.2m / s 2 ; Angular acceleration is limited to 50° / s 2 The parameters of the evaluation function are: α=0.1, β=0.05, γ=0.2, ζ=0.5, and the prediction time period is set to 3.0s.

[0129] The improved trajectory evaluation function is as follows:

[0130]

[0131] Where: αPHead(v,ω) is the angle difference between the end point of the trajectory and the current target point; dist(v,ω) is the distance between the trajectory and the nearest obstacle; vel(v,ω) is the current speed evaluation function; path(v,ω) is the path evaluation subfunction, σ is the smoothing function; α, β, γ are weighting coefficients, θ ob is the angle factor.

[0132] It is best to refine the path between a series of key nodes extracted from the global optimal path based on the improved DWA algorithm, and dynamically adjust the path through the current speed, direction and obstacle information of the quadruped robot, so that the robot can flexibly avoid close obstacles and adjust the motion trajectory in real time to ensure that the quadruped robot moves efficiently and safely along the optimal path.

[0133] Step 5: Motion Control:

[0134] The control module issues control instructions and coordinates the movement of the quadruped robot based on the path and obstacle avoidance information provided by the navigation module;

[0135] Execute navigation instructions to realize autonomous movement of the robot;

[0136] Monitor and adjust the robot's posture and gait to ensure smooth movement.

[0137] The specific steps include:

[0138] Control instructions are issued: The control module receives the path information and obstacle avoidance information provided by the navigation module and the perception module in real time; based on the received information, it calculates the necessary walking instructions, including the adjustment of speed, direction and gait; the control instructions are sent to the quadruped robot drive system through the internal communication protocol;

[0139] Execute navigation instructions: After receiving the control instructions, the drive system drives the joints and motors of the quadruped robot to move according to the specified motion mode; through feedback systems such as encoders and angle sensors, the accuracy and stability of the execution process are monitored to ensure the correct implementation of the instructions;

[0140] Posture and gait adjustment: Monitor the posture data of the quadruped robot in real time and collect it through gyroscope attitude sensors and accelerometers; dynamically adjust the gait and body posture of the quadruped robot based on the monitoring data, and use the MPC algorithm (model predictive control) to fine-tune the joint angle and step frequency to ensure that the quadruped robot can move smoothly on different terrains.

[0141] Step 6: Data Interaction:

[0142] The communication module realizes real-time data interaction between the quadruped robot and the cloud platform and inspection system through WiFi, 4G or 5G network, and supports remote monitoring and control;

[0143] Step 7: Task Execution:

[0144] Perform inspection tasks according to the instructions of the task management unit; monitor the progress of tasks in real time to ensure smooth completion of tasks; record and report data collected during the inspection process.

[0145] The specific steps include:

[0146] Inspection task execution: receiving inspection tasks assigned by the task management unit to the quadruped robot, including the task location, target inspection items and scheduled timetable; breaking down the overall task into several subtasks, inspecting specific equipment and collecting specific environmental data, and arranging the time sequence and priority;

[0147] Task monitoring and adjustment: During the task execution, the position, status and environmental changes of the quadruped robot are monitored in real time; according to the actual situation, if the quadruped robot deviates from the planned path or the task is delayed, the system automatically adjusts the travel path or rearranges the execution order of subtasks;

[0148] Data collection and reporting: When performing inspection tasks, the required data, such as environmental information images, are collected through the sensor system; the collected data is integrated and an inspection report is generated, which is sent to the central control system in real time for analysis and recording.

[0149] Through the above steps, the autonomous navigation and dynamic obstacle avoidance of the intelligent inspection quadruped robot in complex environments can be realized, thus improving the efficiency and safety of inspection tasks.

[0150] In summary, the present invention provides an autonomous navigation method and system for an intelligent inspection quadruped robot. Through multi-sensor data fusion and an improved fusion path planning algorithm, the robot's autonomous navigation capability in the environment is enhanced, the inspection efficiency and safety are improved, and strong technical support is provided for intelligent management.

[0151] Matters not covered in the present invention are known technologies. The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. An autonomous navigation system for an intelligent inspection quadruped robot, characterized in that: Including perception module, navigation module, control module and communication module; The perception module is used for environmental perception, responsible for collecting environmental information and transmitting data to the navigation module for path planning; The navigation module is responsible for path planning and automatic obstacle avoidance. After planning the optimal path based on the perception data, it sends instructions to the control module. The control module is used to execute control instructions and control the movement of the robot according to navigation instructions; at the same time, the communication module realizes interaction with external systems and is responsible for data transmission and communication between the robot and other systems or devices, ensuring the coordinated work between various modules to achieve efficient and accurate inspection tasks.

2. The autonomous navigation system of an intelligent inspection quadruped robot according to claim 1, characterized in that: The perception module includes a laser radar and a depth camera, which are used to obtain surrounding environment information in real time and form an environment map; The laser radar is installed on the top of the quadruped robot to maximize its field of view and detection capability; The depth camera is installed on the face of the robot to capture three-dimensional information of the environment in front.

3. The autonomous navigation system of an intelligent inspection quadruped robot according to claim 1, characterized in that: The navigation module uses the improved DWA algorithm and the A* algorithm to fuse the environment map provided by the perception module, uses the A* algorithm to generate a global path, extracts the optimal node and outputs the global optimal path, and then uses the dynamic window method to perform local path planning between adjacent nodes after obtaining the global optimal node sequence, and performs dynamic obstacle avoidance and path adjustment in a local range to achieve efficient autonomous navigation; An exponential function with a distance attenuation factor is introduced into the obstacle distance evaluation function of the DWA algorithm, as shown below: f(x)=e -δx Among them, x represents the actual distance between the quadruped robot and the obstacle, δ is the distance attenuation coefficient, and the improved obstacle distance evaluation function is as follows: Among them, (x, y) is the coordinate of the starting point of the quadruped robot, (x ob ,y ob ) is the coordinate of the target point of the quadruped robot, r is the safety radius; The improved DWA algorithm introduces a path evaluation sub-function path in the DWA algorithm, and the smaller the weight is, the closer it is to the global path planning, and vice versa, the more it deviates from the global path planning; Then introduce the angle factor θ ob , which is the angle between the quadruped robot and the obstacle, θ ob is always less than 180°, and then θ ob It is added to the denominator of the trajectory evaluation function to dynamically adjust the weight of the sub-evaluation function; The improved trajectory evaluation function is as follows: Where: αPHead(v,ω) is the angle difference between the end point of the trajectory and the current target point; dist(v,ω) is the distance between the trajectory and the nearest obstacle; vel(v,ω) is the current speed evaluation function; path(v,ω) is the path evaluation subfunction, σ is the smoothing function; α, β, γ are weighting coefficients, θ ob is the angle factor.

4. The autonomous navigation system of an intelligent inspection quadruped robot according to claim 1, characterized in that: The control module includes a motion control unit and a task management unit. The motion control unit is responsible for the coordination and control of the robot's quadruped motion. The task management unit issues execution instructions based on the path and obstacle avoidance information provided by the navigation module to achieve autonomous inspection tasks. The communication module adopts wireless communication technology WiFi, 4G or 5G network to realize real-time data interaction between the quadruped robot and the cloud platform and inspection system, and supports remote monitoring and control.

5. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to any one of claims 1 to 4, characterized in that: The steps include: Step 1: Initialize the system: Turn on the quadruped robot and start the perception module, navigation module, control module and communication module; Step 2: Environmental Perception: The perception module acquires environmental data in real time through lidar and depth camera and generates an environmental map; Step 3: Global path planning: The navigation module uses the A* algorithm to generate a global path based on the environment map, extract the optimal node and output the global optimal path to obtain the global optimal node sequence; Step 4: Local path planning and obstacle avoidance: During the local path planning process, the navigation module uses the improved DWA algorithm for dynamic obstacle avoidance and path adjustment; Step 5: Motion Control: The control module issues control instructions based on the path and obstacle avoidance information provided by the navigation module to coordinate the movement of the quadruped robot; executes navigation instructions to achieve autonomous movement of the robot; monitors and adjusts the robot's posture and gait to ensure smooth movement; Step 6: Data Interaction: The communication module realizes real-time data interaction between the quadruped robot and the cloud platform and inspection system through WiFi, 4G or 5G network, and supports remote monitoring and control; Step 7: Task Execution: Perform inspection tasks according to the instructions of the task management unit; Monitor task progress in real time to ensure successful completion of tasks; record and report data collected during the inspection process.

6. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to claim 5, characterized in that: The specific steps of step 2 include: The laser radar on the top of the quadruped robot is used to obtain point cloud data of the surrounding environment, and the point cloud data is subjected to noise reduction and filtering. The depth camera on the face of the quadruped robot is used to obtain the depth image of the surrounding environment, and the image processing algorithm is used to identify obstacles and terrain features in the real-time environment; Project the LiDAR point cloud data onto a two-dimensional plane to generate a two-dimensional point cloud projection map. Then convert the depth image data of the depth camera into corresponding three-dimensional points and project them onto the same two-dimensional plane. Align the point cloud projection map and the depth map projection map, perform data fusion, and ensure the spatial consistency of the two sensor data. According to the predetermined environment size and resolution, an empty two-dimensional grid map is initialized. Each grid represents a small area in the environment, usually a square. White grids represent idle states and black grids represent obstacles. The Bayesian update rule is then used to fuse the lidar and depth camera data into the grid map. The final output two-dimensional grid map will contain obstacle information in the environment for robot navigation and path planning.

7. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to claim 5, characterized in that: The specific steps of step 3 include: Combined with the data from the environment perception module, a two-dimensional grid map of the current environment is generated; Set the robot's starting position and target position according to the task requirements and mark them on the environment map; The A* algorithm is used to start from the starting node, gradually evaluate possible path nodes, calculate f(n) in each step, and select the node with the smallest f(n) for expansion until the target node is reached; The A* algorithm uses Euclidean distance to define the heuristic function, which is: In the formula, n represents the current node of the robot, h(n) represents the distance from the current node to the target node, and n x Indicates the x-axis coordinate of the current node, n y represents the y-axis coordinate of the current node; g represents the robot's target node, g x Indicates the x-axis coordinate of the target node, g y Indicates the y-axis coordinate of the target node; The cost function formula of the A* algorithm is: f(n)=g(n)+h(n) Where f(n) represents the estimated total cost from the starting node to the target node via node n, g(n) represents the actual distance from the starting node to node n in the state space, and h(n) represents the estimated cost from node n to the target node; Finally, the node sequence in the expansion process is recorded, and the node sequence of the optimal path is finally output.

8. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to claim 7, characterized in that: The specific steps of step 4 include: An exponential function with a distance attenuation factor is introduced into the obstacle distance evaluation function of the DWA algorithm, as shown below: f(x)=e -δx In the formula, x represents the actual distance between the quadruped robot and the obstacle, δ is the distance attenuation coefficient, and the improved obstacle distance evaluation function is as follows: Among them, (x, y) is the coordinate of the starting point of the quadruped robot, (x ob ,y ob ) is the coordinate of the target point of the quadruped robot, r is the safety radius. From the formula, we know that the size of δ can adjust the degree of attention to different obstacles in path planning; The improved DWA algorithm introduces a path evaluation sub-function path in the DWA algorithm, and the smaller the weight, the closer it is to the global path planning, and vice versa, the more it deviates from the global path planning; Then introduce the angle factor θ ob , which is the angle between the quadruped robot and the obstacle, θ ob is always less than 180°, and then θ ob It is added to the denominator of the trajectory evaluation function to dynamically adjust the weight of the sub-evaluation function; The improved trajectory evaluation function is as follows: Where: αPHead(v,ω) is the angle difference between the end point of the trajectory and the current target point; dist(v,ω) is the distance between the trajectory and the nearest obstacle; vel(v,ω) is the current speed evaluation function; path(v,ω) is the path evaluation subfunction, σ is the smoothing function; α, β, γ are weighting coefficients, θ ob is the angle factor.

9. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to claim 5, characterized in that: The step 5 specifically includes: Control instructions are issued: The control module receives the path information and obstacle avoidance information provided by the navigation module and the perception module in real time; based on the received information, it calculates the necessary walking instructions, including the adjustment of speed, direction and gait; the control instructions are sent to the quadruped robot drive system through the internal communication protocol; Execute navigation instructions: After receiving the control instructions, the drive system drives the joints and motors of the quadruped robot to move according to the specified motion mode; through feedback systems such as encoders and angle sensors, the accuracy and stability of the execution process are monitored to ensure the correct implementation of the instructions; Posture and gait adjustment: Monitor the posture data of the quadruped robot in real time and collect it through gyroscope attitude sensors and accelerometers; dynamically adjust the gait and body posture of the quadruped robot based on the monitoring data, and use the MPC algorithm to fine-tune the joint angle and step frequency to ensure that the quadruped robot can move smoothly on different terrains.

10. The method for using the autonomous navigation system of the intelligent inspection quadruped robot according to claim 5, characterized in that: The step 7 specifically includes: Inspection task execution: receiving inspection tasks assigned by the task management unit to the quadruped robot, including the task location, target inspection items and scheduled timetable; breaking down the overall task into several subtasks, inspecting specific equipment and collecting specific environmental data, and arranging the time sequence and priority; Task monitoring and adjustment: During the task execution, the position, status and environmental changes of the quadruped robot are monitored in real time; according to the actual situation, if the quadruped robot deviates from the planned path or the task is delayed, the system automatically adjusts the travel path or rearranges the execution order of subtasks; Data collection and reporting: When performing inspection tasks, the required data, such as environmental information images, are collected through the sensor system; the collected data is integrated and an inspection report is generated, which is sent to the central control system in real time for analysis and recording.

Images