An anti-explosion area autonomous inspection method and system for a robot dog

Abstract

The application discloses an anti-explosion area autonomous inspection method and system for a robot dog, and relates to the technical field of robot intelligent inspection.A kind of anti-explosion area autonomous inspection system for robot dog, including map construction and sequencing module, global path planning module, global path smoothing module, local path optimization module and state detection and alarm module.The application carries out the optimal inspection path planning of two-dimensional space between two adjacent starting points or inspection points by improved A* algorithm, which can not only dynamically adjust the inspection path according to the width of the channel, that is, preferentially select wider channels to effectively reduce the risk of collision, and ensure that the robot dog maintains a safe distance from the channel wall in narrow channels, thereby improving the safety and reliability of the robot dog inspection, but also can improve the passability and inspection efficiency of the inspection path by combining the kinematic characteristics of the robot dog, to reduce unnecessary detours, and improve the safety of the anti-explosion area autonomous inspection method and system for the robot dog.

Description

Technical Field

[0001] This invention relates to the field of intelligent machine inspection technology, specifically to a method and system for autonomous inspection of explosion-proof areas using a robot dog. Background Technology

[0002] Explosion-proof areas refer to areas where explosive gas or flammable dust environments may occur during production processes, such as oil extraction platforms, chemical workshops, gas stations, and underground coal mines. These areas contain flammable and explosive gases, liquids, dust, or fibers, which can easily trigger explosions due to static electricity, electrical sparks, and high temperatures, posing a serious threat to personnel safety and equipment. Therefore, in order to promptly detect and address potential safety hazards in explosion-proof areas, such as equipment leaks, electrical faults, and static electricity accumulation, and to prevent them from developing into fires or explosions that could cause injury or death to personnel and equipment, relevant companies need to conduct regular inspections of explosion-proof areas. To reduce the safety risks of inspections and improve efficiency, companies typically use robots or robotic dogs to replace manual inspections of explosion-proof areas.

[0003] Existing robots or robot dogs typically determine the inspection sequence based on the locations of multiple inspection points when inspecting explosion-proof areas. Then, they plan a global path between two adjacent inspection points based on the obstacle distribution between them. Finally, during the inspection, various sensors and devices monitor the position of dynamic obstacles and dynamically adjust the local path accordingly. However, when there are many dynamic obstacles and they move at high speeds, robots or robot dogs are prone to slow reactions and obstacle avoidance failures. Moreover, conventional inspection path planning methods lack consideration for path width and the stability of the robot or robot dog on the path, which may cause the robot or robot dog to tip over or even collide with surrounding obstacles during the inspection process. As a result, the safety of existing autonomous inspection methods and systems for explosion-proof areas is relatively low.

[0004] Based on the above, this invention proposes a highly secure autonomous inspection method and system for explosion-proof areas used in robot dogs. Summary of the Invention

[0005] To overcome the shortcomings of existing robots or robot dogs in inspecting explosion-proof areas, such as slow reaction and obstacle avoidance failure when encountering a large number of dynamic obstacles at high speeds, and the lack of consideration for path width and robot or robot dog stability in conventional inspection path planning methods, which may cause the robot or robot dog to tip over or collide with surrounding obstacles during inspection, resulting in low safety of existing autonomous inspection methods and systems for explosion-proof areas, this invention proposes a highly safe autonomous inspection method and system for explosion-proof areas using robot dogs.

[0006] A method for autonomous inspection of explosion-proof areas using a robot dog includes the following steps:

[0007] The explosion-proof area is scanned and data is collected in all directions by LiDAR, camera and sensor to obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map. The location information of the starting point and multiple inspection points is obtained. The inspection order and original path of the robot dog are obtained by ant colony algorithm.

[0008] For any two adjacent starting points or inspection points in the inspection sequence of the robot dog, the improved A* algorithm is used to plan the optimal inspection path in two-dimensional space to obtain the global inspection path between the two adjacent starting points or inspection points.

[0009] Based on the improved B-spline curve fitting, the global inspection path between all two adjacent starting points or inspection points is smoothed to obtain the global smooth inspection path between two adjacent starting points or inspection points.

[0010] The explosion-proof area is scanned in real time to obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. The future state information of the dynamic obstacles is predicted by a motion prediction model based on convolutional neural network. Based on the prediction results and combined with the robot dog's own motion state, the model prediction control algorithm is used to optimize the local obstacle avoidance path.

[0011] The system collects data on the environment and equipment around the inspection points using sensors and cameras, analyzes and processes the collected data, and uses fault diagnosis models to assess the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

[0012] As a preferred aspect of the invention, the two-dimensional planar information includes the location information of static obstacles, the width information of the passage, and the equipment layout information within the explosion-proof area, and the future state information includes the location information, direction of movement information, speed of movement information, and acceleration information of dynamic obstacles.

[0013] As a preferred aspect of the invention, the specific steps for using the improved A* algorithm to plan the optimal inspection path in two-dimensional space and obtain the global inspection path between two adjacent starting points or inspection points are as follows:

[0014] The two-dimensional map of the explosion-proof area is discretized into a grid, where each grid represents a node in the explosion-proof area. Based on the two-dimensional planar information of the explosion-proof area, the state of each node is divided into unobstructed or obstructed.

[0015] Two adjacent starting points or inspection points are used as the starting point and target point of the robot dog, respectively, and their position coordinates are determined. The open list and the closed list are initialized, and the starting point is added to the open list. The open list is used to store nodes to be evaluated, and the closed list is used to store nodes that have been evaluated. The heuristic function is set to the Manhattan distance calculation formula.

[0016] Define a heuristic cost function ,in This represents the actual cost from the starting point to the current node. This represents the heuristic cost estimate from the current node to the target node. This indicates the channel width constraint term, and , Indicates the channel width. Indicates the weighting coefficient of the channel width constraint term;

[0017] Remove the node with the minimum heuristic cost from the open list. Generate nodes For each neighboring node Determine the node If the state is obstructed, discard it; otherwise, check the node. Check if the node is in the open or closed list. If it is already in the closed list, skip it; if it is neither in the closed nor open list, add it to the open list and update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ;

[0018] If node If already in the open list, determine whether to proceed through the current node. Reaching the adjacent node Is the path better? If so, update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ;

[0019] The current node Add the node to the closed list and remove it from the open list. Continue to select the next node with the lowest heuristic cost from the open list for evaluation until the target node is selected.

[0020] After finding the target point, the system traces back to the starting point by backtracking the parent node information of each node, constructs and outputs the global inspection path between the two adjacent starting points or inspection points.

[0021] As a preferred aspect of the invention, the specific steps for smoothing the global inspection path between all adjacent starting points or inspection points based on improved B-spline curve fitting to obtain a globally smoothed inspection path between adjacent starting points or inspection points are as follows:

[0022] Obtain the robot dog's size parameters, including length and width, and set the minimum safe distance between the robot dog and static obstacles based on the robot dog's size and inspection task requirements. Control points are selected evenly from the global inspection path, and the start and end points of the global inspection path are selected as control points.

[0023] For each control point Calculate the minimum distance between it and the static obstacle. ,if Then move the control point outwards so that the distance between it and the static obstacle meets the following conditions. The direction of movement is either away from static obstacles or outward along the normal of the global inspection path. The control point after movement... The specific positional expression is:

[0024]

[0025] in Control point The normal direction vector at that point, using the moved control points. Replace control points ;

[0026] Selected control points These serve as control points for a cubic B-spline curve, where the mathematical expression for the cubic B-spline curve is:

[0027]

[0028] in Describes the cubic B-spline basis functions. Indicates the total number of control points;

[0029] Through cubic B-spline curves The formula calculates the points on the B-spline curve and forms a smooth curve. Using the smooth curve as a reference, the robot dog's maximum width is expanded to both sides to form a detection curve. The sign distance between the points on the detection curve and the static obstacle is calculated. If there is a case where it is less than zero, the position of the corresponding control point is adjusted and the B-spline curve is regenerated. This process is repeated until the generated B-spline curve meets the detection requirements, thus obtaining a globally smooth inspection path between two adjacent starting points or inspection points.

[0030] As a preferred aspect of the invention, the specific steps of optimizing the local obstacle avoidance path using a model predictive control algorithm based on the prediction results and the robot dog's own motion state are as follows:

[0031] Obtain the robot dog's current motion state information, including position, speed, acceleration, and orientation; set MPC parameters, including prediction time domain length, control time domain length, and sampling time.

[0032] Construct an optimization problem, define state variables and control variables, where the state variables include the robot dog's position, velocity, acceleration and orientation, and the control variables include the robot dog's acceleration and angular velocity. Based on the robot dog's kinematic model, construct the state transition equation.

[0033] The objective function is defined based on safety distance, inspection efficiency, and motion stability. The specific expression of the objective function is as follows:

[0034]

[0035] in Indicates the time relationship between the robot dog and dynamic obstacles. Predicted safe distance, Indicates the robot dog in time speed, Indicates the robot dog in time acceleration, , and These are the weighting coefficients for safe distance, inspection efficiency, and motion stability, respectively.

[0036] Set constraints, including safety distance constraints, robot dog dynamics constraints, and path feasibility constraints;

[0037] In each control cycle, an optimization problem is solved based on the robot dog's current motion state and the future state information of dynamic obstacles to obtain the optimal control output. The optimal control input is then applied to the robot dog, and the robot dog's motion state is updated in real time to achieve dynamic obstacle avoidance and obtain an optimized local obstacle avoidance path.

[0038] As a preferred aspect of the invention, the safety distance constraint is ,in This represents the preset minimum safe distance; the robot dog's dynamic constraints are... , and ,in , and These represent the robot dog's minimum speed, minimum acceleration, and minimum turning angle, respectively. , and These represent the robot dog's maximum speed, maximum acceleration, and maximum turning angle, respectively. The path feasibility constraint requires that the optimized local obstacle avoidance path and the global smooth inspection path maintain continuity within a preset range, wherein continuity is quantified and constrained by path angle and curvature.

[0039] An autonomous inspection system for explosion-proof areas for robotic dogs, comprising:

[0040] The map building and sorting module is used to perform all-round scanning and data collection of the explosion-proof area through LiDAR, camera and sensor, obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map, obtain the location information of the starting point and multiple inspection points, and obtain the inspection order and original path of the robot dog through ant colony algorithm.

[0041] The global path planning module is used to plan the optimal inspection path in two-dimensional space for any two adjacent starting points or inspection points in the inspection sequence of the robot dog, using an improved A* algorithm, to obtain the global inspection path between the two adjacent starting points or inspection points.

[0042] The global path smoothing module is used to smooth the global inspection path between all two adjacent starting points or inspection points based on the improved B-spline curve fitting, so as to obtain the global smooth inspection path between two adjacent starting points or inspection points.

[0043] The local path optimization module is used to perform a full-range real-time scan of the explosion-proof area, obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. It uses a motion prediction model based on a convolutional neural network to predict the future state information of dynamic obstacles. Based on the prediction results and combined with the robot dog's own motion state, it uses a model prediction control algorithm to optimize the local obstacle avoidance path.

[0044] The status detection and alarm module is used to collect data on the environment and equipment around the inspection point through sensors and cameras, analyze and process the collected data, and use a fault diagnosis model to evaluate the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

[0045] The present invention has the following advantages:

[0046] 1. This invention uses an improved A* algorithm to plan the optimal inspection path in two-dimensional space between two adjacent starting points or inspection points. It can not only dynamically adjust the inspection path according to the channel width, i.e., prioritize wider channels to effectively reduce the risk of collision, and ensure that the robot dog maintains a safe distance from the channel wall in narrow channels, thereby improving the safety and reliability of robot dog inspection, but also improve the passability and inspection efficiency of the inspection path by combining the kinematic characteristics of the robot dog, so as to reduce unnecessary detours and improve the safety of this autonomous inspection method and system for explosion-proof areas of robot dogs.

[0047] 2. This invention uses improved B-spline curve fitting to smooth the global inspection path between all adjacent starting points or inspection points. This not only makes the global inspection path smoother, thereby improving the fluency and stability of the robot dog's inspection, but also adjusts the global inspection path based on the robot dog's length and width and the path width. This effectively avoids collisions between the robot dog and obstacles during inspection, ensuring the smooth completion of the inspection task and improving the safety of this autonomous inspection method and system for explosion-proof areas.

[0048] 3. This invention predicts the future state information of dynamic obstacles using a motion prediction model based on convolutional neural networks, enabling the early acquisition of obstacle movement trends and providing prior information for the robot dog's local obstacle avoidance. By combining the prediction results with the robot dog's own motion state and using a model predictive control algorithm to optimize the local obstacle avoidance path, dynamic obstacles can be effectively avoided while ensuring a safe distance. This also takes into account inspection efficiency and motion stability. The algorithm's objective function comprehensively considers safe distance, inspection efficiency, and motion stability, allowing the robot dog to quickly and smoothly adjust its path in dynamic environments to avoid collisions with dynamic obstacles. This improves the execution efficiency and reliability of inspection tasks, enhancing the safety and practicality of this autonomous inspection method and system for explosion-proof areas used in robot dogs. Attached Figure Description

[0049] Figure 1 This is a flowchart illustrating a method for autonomous inspection of explosion-proof areas using a robot dog, as described in an embodiment of the present invention.

[0050] Figure 2 This is a schematic diagram of the structure of an explosion-proof autonomous inspection system for a robot dog, as used in an embodiment of the present invention. Detailed Implementation

[0051] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention.

[0052] Example 1: A method for autonomous inspection of explosion-proof areas using a robot dog, such as... Figure 1 As shown, it includes the following steps:

[0053] The explosion-proof area is scanned and data is collected in all directions by LiDAR, camera and sensor to obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map. The location information of the starting point and multiple inspection points is obtained. The inspection order and original path of the robot dog are obtained by ant colony algorithm.

[0054] For any two adjacent starting points or inspection points in the inspection sequence of the robot dog, the improved A* algorithm is used to plan the optimal inspection path in two-dimensional space to obtain the global inspection path between the two adjacent starting points or inspection points.

[0055] Based on the improved B-spline curve fitting, the global inspection path between all two adjacent starting points or inspection points is smoothed to obtain the global smooth inspection path between two adjacent starting points or inspection points.

[0056] The explosion-proof area is scanned in real time to obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. The future state information of the dynamic obstacles is predicted by a motion prediction model based on convolutional neural network. Based on the prediction results and combined with the robot dog's own motion state, the model prediction control algorithm is used to optimize the local obstacle avoidance path.

[0057] The system collects data on the environment and equipment around the inspection points using sensors and cameras, analyzes and processes the collected data, and uses fault diagnosis models to assess the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

[0058] It should be noted that obtaining the robot dog's inspection sequence and original path based on the location information of the starting point and multiple inspection points using the ant colony algorithm is a mature existing technology, and therefore will not be elaborated upon here. However, predicting the future state information of dynamic obstacles using a motion prediction model based on a convolutional neural network requires prior collection of a large amount of motion trajectory data of high-speed dynamic obstacles (such as forklifts, automated guided vehicles, etc.) in different scenarios, including information such as position, speed, acceleration, and turning angle, as training samples to train the motion prediction model. In actual inspections, the robot dog uses the motion prediction model to extract and analyze the features of the instantaneous motion state of high-speed dynamic obstacles, and can predict their position and speed change trends in the next few seconds. The collected data is then analyzed and processed, and a fault diagnosis model is applied. The assessment of equipment operating status includes, but is not limited to, combining gas concentration data with temperature and humidity data through data fusion algorithms (such as Bayesian fusion algorithm and Kalman filter fusion algorithm), considering the influence of ambient temperature and humidity on the gas sensor measurement results, correcting the gas concentration data to obtain a more accurate actual gas concentration value, and issuing an alarm when the actual gas concentration value exceeds a preset threshold. It also involves spatially registering and fusing image data with infrared thermal imaging data, and overlaying temperature distribution information onto the equipment's external image to generate an intuitive visualization of the equipment's status. This allows remote monitoring personnel to quickly and accurately determine the equipment's operating status. For example, remote monitoring personnel can clearly see an abnormal temperature rise in a certain part of the equipment and accurately locate that part on the image, thereby promptly identifying potential fault points and issuing alarms.

[0059] The two-dimensional planar information includes the location information of static obstacles, the width information of the passage, and the equipment layout information within the explosion-proof area. The future state information includes the location information, direction of movement information, speed of movement information, and acceleration information of dynamic obstacles.

[0060] The specific steps for using the improved A* algorithm to plan the optimal inspection path in two-dimensional space and obtain the global inspection path between two adjacent starting points or inspection points are as follows:

[0061] The two-dimensional map of the explosion-proof area is discretized into a grid, where each grid represents a node in the explosion-proof area. Based on the two-dimensional planar information of the explosion-proof area, the state of each node is divided into unobstructed or obstructed.

[0062] Two adjacent starting points or inspection points are used as the starting point and target point of the robot dog, respectively, and their position coordinates are determined. The open list and the closed list are initialized, and the starting point is added to the open list. The open list is used to store nodes to be evaluated, and the closed list is used to store nodes that have been evaluated. The heuristic function is set to the Manhattan distance calculation formula.

[0063] Define a heuristic cost function ,in This represents the actual cost from the starting point to the current node. This represents the heuristically estimated cost from the current node to the target point, i.e., the Manhattan distance from the current node to the target point. This indicates the channel width constraint term, and , Indicates the channel width. Indicates the weighting coefficient of the channel width constraint term;

[0064] Remove the node with the minimum heuristic cost from the open list. Generate nodes For each neighboring node Determine the node If the state is obstructed, discard it; otherwise, check the node. Check if the node is in the open or closed list. If it is already in the closed list, skip it; if it is neither in the closed nor open list, add it to the open list and update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ;

[0065] If node If already in the open list, determine whether to proceed through the current node. Reaching the adjacent node Is the path better? If so, update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ;

[0066] The current node Add the node to the closed list and remove it from the open list. Continue to select the next node with the lowest heuristic cost from the open list for evaluation until the target node is selected.

[0067] After finding the target point, the system traces back to the starting point by backtracking the parent node information of each node, constructs and outputs the global inspection path between the two adjacent starting points or inspection points.

[0068] It should be noted that when planning the global inspection path, high-density obstacle areas (i.e., areas where the number of obstacles exceeds a preset threshold) can be identified first based on the obstacle density information in the map. Then, the status of all nodes contained in the entire area is considered to be obstructed, so that these areas can be avoided from a macro perspective during path planning, thereby effectively simplifying the global inspection path.

[0069] The above steps utilize an improved A* algorithm to plan the optimal inspection path in two-dimensional space between two adjacent starting points or inspection points. This not only dynamically adjusts the inspection path based on the channel width (prioritizing wider channels to effectively reduce collision risks and ensuring the robot dog maintains a safe distance from the channel wall in narrow channels, thereby improving the safety and reliability of the robot dog's inspection), but also enhances the passability and efficiency of the inspection path by combining the robot dog's kinematic characteristics, reducing unnecessary detours and improving the safety of this autonomous inspection method and system for explosion-proof areas.

[0070] The specific steps for smoothing the global inspection path between all adjacent starting points or inspection points based on improved B-spline curve fitting to obtain a globally smoothed inspection path between adjacent starting points or inspection points are as follows:

[0071] Obtain the robot dog's size parameters, including length and width, and set the minimum safe distance between the robot dog and static obstacles based on the robot dog's size and inspection task requirements. Control points are selected evenly from the global inspection path, and the start and end points of the global inspection path are selected as control points.

[0072] For each control point Calculate the minimum distance between it and the static obstacle. ,if Then move the control point outwards so that the distance between it and the static obstacle meets the following conditions. The direction of movement is either away from static obstacles or outward along the normal of the global inspection path. The control point after movement... The specific positional expression is:

[0073]

[0074] in Control point The normal direction vector at that point, using the moved control points. Replace control points ;

[0075] Selected control points These serve as control points for a cubic B-spline curve, where the mathematical expression for the cubic B-spline curve is:

[0076]

[0077] in Describes the cubic B-spline basis functions. Indicates the total number of control points;

[0078] Through cubic B-spline curves The formula calculates the points on the B-spline curve and forms a smooth curve. Using the smooth curve as a reference, the robot dog's maximum width is expanded to both sides to form a detection curve. The sign distance between the points on the detection curve and the static obstacle is calculated. If there is a case where it is less than zero, the position of the corresponding control point is adjusted and the B-spline curve is regenerated. This process is repeated until the generated B-spline curve meets the detection requirements, thus obtaining a globally smooth inspection path between two adjacent starting points or inspection points.

[0079] The above steps smooth the global inspection path between all adjacent starting points or inspection points using improved B-spline curve fitting. This not only makes the global inspection path smoother, thereby improving the fluency and stability of the robot dog's inspection, but also adjusts the global inspection path based on the robot dog's length and width and the path width. This effectively avoids collisions between the robot dog and obstacles during inspection, ensuring the smooth completion of the inspection task and improving the safety of this autonomous inspection method and system for explosion-proof areas.

[0080] The specific steps for optimizing the local obstacle avoidance path using a model predictive control algorithm based on the prediction results and the robot dog's own motion state are as follows:

[0081] Obtain the robot dog's current motion state information, including position, speed, acceleration, and orientation; set MPC parameters, including prediction time domain length, control time domain length, and sampling time.

[0082] Construct an optimization problem, define state variables and control variables, where the state variables include the robot dog's position, velocity, acceleration and orientation, and the control variables include the robot dog's acceleration and angular velocity. Based on the robot dog's kinematic model, construct the state transition equation.

[0083] The objective function is defined based on safety distance, inspection efficiency, and motion stability. The specific expression of the objective function is as follows:

[0084]

[0085] in Indicates the time relationship between the robot dog and dynamic obstacles. Predicted safe distance, Indicates the robot dog in time speed, Indicates the robot dog in time acceleration, , and These are the weighting coefficients for safe distance, inspection efficiency, and motion stability, respectively.

[0086] Set constraints, including safety distance constraints, robot dog dynamics constraints, and path feasibility constraints;

[0087] In each control cycle, an optimization problem is solved based on the robot dog's current motion state and the future state information of dynamic obstacles to obtain the optimal control output. The optimal control input is then applied to the robot dog, and the robot dog's motion state is updated in real time to achieve dynamic obstacle avoidance and obtain an optimized local obstacle avoidance path.

[0088] The safety distance constraint is as follows: ,in This represents the preset minimum safe distance; the robot dog's dynamic constraints are... , and ,in , and These represent the robot dog's minimum speed, minimum acceleration, and minimum turning angle, respectively. , and These represent the robot dog's maximum speed, maximum acceleration, and maximum turning angle, respectively. The path feasibility constraint requires that the optimized local obstacle avoidance path and the global smooth inspection path maintain continuity within a preset range, wherein continuity is quantified and constrained by path angle and curvature.

[0089] The above steps predict the future state information of dynamic obstacles using a motion prediction model based on convolutional neural networks, enabling the early acquisition of obstacle movement trends and providing prior information for the robot dog's local obstacle avoidance. By combining the prediction results with the robot dog's own motion state and using a model predictive control algorithm to optimize the local obstacle avoidance path, dynamic obstacles can be effectively avoided while ensuring a safe distance. This also takes into account inspection efficiency and motion stability. The algorithm's objective function comprehensively considers safe distance, inspection efficiency, and motion stability, enabling the robot dog to quickly and smoothly adjust its path in dynamic environments to avoid collisions with dynamic obstacles. This improves the execution efficiency and reliability of inspection tasks, enhancing the safety and practicality of this autonomous inspection method and system for explosion-proof areas used by robot dogs.

[0090] Example 2: An autonomous inspection system for explosion-proof areas for robot dogs, such as... Figure 2 As shown, it includes:

[0091] The map building and sorting module is used to perform all-round scanning and data collection of the explosion-proof area through LiDAR, camera and sensor, obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map, obtain the location information of the starting point and multiple inspection points, and obtain the inspection order and original path of the robot dog through ant colony algorithm.

[0092] The global path planning module is used to plan the optimal inspection path in two-dimensional space for any two adjacent starting points or inspection points in the inspection sequence of the robot dog, using an improved A* algorithm, to obtain the global inspection path between the two adjacent starting points or inspection points.

[0093] The global path smoothing module is used to smooth the global inspection path between all two adjacent starting points or inspection points based on the improved B-spline curve fitting, so as to obtain the global smooth inspection path between two adjacent starting points or inspection points.

[0094] The local path optimization module is used to perform a full-range real-time scan of the explosion-proof area, obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. It uses a motion prediction model based on a convolutional neural network to predict the future state information of dynamic obstacles. Based on the prediction results and combined with the robot dog's own motion state, it uses a model prediction control algorithm to optimize the local obstacle avoidance path.

[0095] The status detection and alarm module is used to collect data on the environment and equipment around the inspection point through sensors and cameras, analyze and process the collected data, and use a fault diagnosis model to evaluate the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

[0096] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A method for autonomous inspection of explosion-proof areas using a robot dog, characterized in that, Includes the following steps: The explosion-proof area is scanned and data is collected in all directions by LiDAR, camera and sensor to obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map. The location information of the starting point and multiple inspection points is obtained. The inspection order and original path of the robot dog are obtained by ant colony algorithm. For any two adjacent starting points or inspection points in the robot dog's inspection sequence, the improved A* algorithm is used to plan the optimal inspection path in two-dimensional space to obtain the global inspection path between the two adjacent starting points or inspection points. The specific steps are as follows: The two-dimensional map of the explosion-proof area is discretized into a grid, where each grid represents a node in the explosion-proof area. Based on the two-dimensional planar information of the explosion-proof area, the state of each node is divided into unobstructed or obstructed. Two adjacent starting points or inspection points are used as the starting point and target point of the robot dog, respectively, and their position coordinates are determined. The open list and the closed list are initialized, and the starting point is added to the open list. The open list is used to store nodes to be evaluated, and the closed list is used to store nodes that have been evaluated. The heuristic function is set to the Manhattan distance calculation formula. Define a heuristic cost function ,in This represents the actual cost from the starting point to the current node. This represents the heuristic cost estimate from the current node to the target node. This indicates the channel width constraint term, and , Indicates the channel width. Indicates the weighting coefficient of the channel width constraint term; Remove the node with the minimum heuristic cost from the open list. Generate nodes For each neighboring node Determine the node If the state is obstructed, discard it; otherwise, check the node. Check if the node is in the open or closed list. If it is already in the closed list, skip it; if it is neither in the closed nor open list, add it to the open list and update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ; If node If already in the open list, determine whether to proceed through the current node. Reaching the adjacent node Is the path better? If so, update the node. The actual cost Heuristic cost estimation Channel width constraint and heuristic costs And record its parent node ; The current node Add the node to the closed list and remove it from the open list. Continue to select the next node with the lowest heuristic cost from the open list for evaluation until the target node is selected. After finding the target point, trace back from the target point to the starting point by backtracking the parent node information of each node, construct and output the global inspection path between the two adjacent starting points or inspection points; Based on the improved B-spline curve fitting, the global inspection path between all two adjacent starting points or inspection points is smoothed to obtain the global smooth inspection path between two adjacent starting points or inspection points. The explosion-proof area is scanned in real time to obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. The future state information of the dynamic obstacles is predicted by a motion prediction model based on convolutional neural network. Based on the prediction results and combined with the robot dog's own motion state, the model prediction control algorithm is used to optimize the local obstacle avoidance path. The system collects data on the environment and equipment around the inspection points using sensors and cameras, analyzes and processes the collected data, and uses fault diagnosis models to assess the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

2. The method for autonomous inspection of explosion-proof areas for a robot dog according to claim 1, characterized in that, The two-dimensional planar information includes the location information of static obstacles, the width information of the passage, and the equipment layout information within the explosion-proof area. The future state information includes the location information, direction of movement information, speed of movement information, and acceleration information of dynamic obstacles.

3. The method for autonomous inspection of explosion-proof areas for a robot dog according to claim 2, characterized in that, The specific steps for smoothing the global inspection path between all adjacent starting points or inspection points based on improved B-spline curve fitting to obtain a globally smoothed inspection path between adjacent starting points or inspection points are as follows: Obtain the robot dog's size parameters, including length and width, and set the minimum safe distance between the robot dog and static obstacles based on the robot dog's size and inspection task requirements. Control points are selected evenly from the global inspection path, and the start and end points of the global inspection path are selected as control points. For each control point Calculate the minimum distance between it and the static obstacle. ,if Then move the control point outwards so that the distance between it and the static obstacle meets the following conditions. The direction of movement is either away from static obstacles or outward along the normal of the global inspection path. The control point after movement... The specific positional expression is: ； in Control point The normal direction vector at that point, using the moved control points. Replace control points ; Selected control points These serve as control points for a cubic B-spline curve, where the mathematical expression for the cubic B-spline curve is: ； in Describes the cubic B-spline basis functions. Indicates the total number of control points; Through cubic B-spline curves The formula calculates the points on the B-spline curve and forms a smooth curve. Using the smooth curve as a reference, the robot dog's maximum width is expanded to both sides to form a detection curve. The sign distance between the points on the detection curve and the static obstacle is calculated. If there is a case where it is less than zero, the position of the corresponding control point is adjusted and the B-spline curve is regenerated. This process is repeated until the generated B-spline curve meets the detection requirements, thus obtaining a globally smooth inspection path between two adjacent starting points or inspection points.

4. The method for autonomous inspection of explosion-proof areas for a robot dog according to claim 3, characterized in that, The specific steps for optimizing the local obstacle avoidance path using a model predictive control algorithm based on the prediction results and the robot dog's own motion state are as follows: Obtain the robot dog's current motion state information, including position, speed, acceleration, and orientation; set MPC parameters, including prediction time domain length, control time domain length, and sampling time. Construct an optimization problem, define state variables and control variables, where the state variables include the robot dog's position, velocity, acceleration and orientation, and the control variables include the robot dog's acceleration and angular velocity. Based on the robot dog's kinematic model, construct the state transition equation. The objective function is defined based on safety distance, inspection efficiency, and motion stability. The specific expression of the objective function is as follows: ； in Indicates the time relationship between the robot dog and dynamic obstacles. Predicted safe distance, Indicates the robot dog in time speed, Indicates the robot dog in time acceleration, , and These are the weighting coefficients for safe distance, inspection efficiency, and motion stability, respectively. Set constraints, including safety distance constraints, robot dog dynamics constraints, and path feasibility constraints; In each control cycle, an optimization problem is solved based on the robot dog's current motion state and the future state information of dynamic obstacles to obtain the optimal control output. The optimal control input is then applied to the robot dog, and the robot dog's motion state is updated in real time to achieve dynamic obstacle avoidance and obtain an optimized local obstacle avoidance path.

5. The method for autonomous inspection of explosion-proof areas for a robot dog according to claim 4, characterized in that, The safety distance constraint is as follows: ,in This represents the preset minimum safe distance; the robot dog's dynamic constraints are... , and ,in , and These represent the robot dog's minimum speed, minimum acceleration, and minimum turning angle, respectively. , and These represent the robot dog's maximum speed, maximum acceleration, and maximum turning angle, respectively. The path feasibility constraint requires that the optimized local obstacle avoidance path and the global smooth inspection path maintain continuity within a preset range, wherein continuity is quantified and constrained by path angle and curvature.

6. An autonomous inspection system for explosion-proof areas for robot dogs, applied to the autonomous inspection method for explosion-proof areas for robot dogs as described in any one of claims 1-5, characterized in that, Including: The map building and sorting module is used to perform all-round scanning and data collection of the explosion-proof area through LiDAR, camera and sensor, obtain two-dimensional planar information of the explosion-proof area and build a corresponding two-dimensional map, obtain the location information of the starting point and multiple inspection points, and obtain the inspection order and original path of the robot dog through ant colony algorithm. The global path planning module is used to plan the optimal inspection path in two-dimensional space for any two adjacent starting points or inspection points in the inspection sequence of the robot dog, using an improved A* algorithm, to obtain the global inspection path between the two adjacent starting points or inspection points. The global path smoothing module is used to smooth the global inspection path between all two adjacent starting points or inspection points based on the improved B-spline curve fitting, so as to obtain the global smooth inspection path between two adjacent starting points or inspection points. The local path optimization module is used to perform a full-range real-time scan of the explosion-proof area, obtain the real-time location of dynamic obstacles in the explosion-proof area and update the two-dimensional map in real time. It uses a motion prediction model based on a convolutional neural network to predict the future state information of dynamic obstacles. Based on the prediction results and combined with the robot dog's own motion state, it uses a model prediction control algorithm to optimize the local obstacle avoidance path. The status detection and alarm module is used to collect data on the environment and equipment around the inspection point through sensors and cameras, analyze and process the collected data, and use a fault diagnosis model to evaluate the operating status of the equipment. When the equipment is in an abnormal state, an alarm is issued.

Images