A method and device for locating a dangerous chemical leakage source of a multi-robot inspection system

Abstract

The application discloses a dangerous chemical leakage source positioning method and device of a multi-robot inspection system, and belongs to the technical field of robot positioning. The method controls each inspection robot to perform cooperative inspection in an obstacle-avoiding formation mode based on autonomous navigation. The inspection robots comprise a leading robot, the leading robot leads the following inspection robots, and the remaining inspection robots follow the preceding inspection robots. When any inspection robot detects a dangerous chemical leakage source, a sound alarm is sent, and gas anomaly data is sent to the leading robot through networking, so that a positioning robot is controlled to search for the dangerous chemical leakage source and accurately position the dangerous chemical leakage source. All formations are based on the following behavior between two robots, and the dependence of the remaining inspection robots on the leading robot in the formation is greatly reduced under the condition that the formation can be maintained, the robustness of cooperative inspection is improved, and the accuracy of positioning is finally ensured.

Description

Technical Field

[0001] This invention belongs to the field of robot positioning technology, and more specifically, relates to a method and device for locating hazardous chemical leak sources in a multi-robot inspection system. Background Technology

[0002] With the rapid pace of industrialization and urbanization, an increasing number of chemicals are being used in society, bringing numerous conveniences to human production and daily life. However, due to factors such as equipment damage and aging, and environmental changes, hazardous chemical leaks have become a ticking time bomb in human production operations. During a leak, toxic and even flammable and explosive gases are released, leading to risks such as poisoning and explosions. Because hazardous chemical inspection is a complex task requiring a certain level of efficiency, and leak sites often involve complex terrain and large areas, robots operating alone struggle to ensure accurate and efficient task completion, resulting in low inspection precision.

[0003] Multiple robots working in a swarm utilize various collaborative technologies, primarily robot formation. In multi-robot inspection systems, robot formation facilitates overall robot movement and improves control efficiency by ensuring smooth scheduling during inspections. Among these, leader-follower formation is the most common: a leader robot guides the swarm, with other robots acting as followers. The leader moves according to task requirements, and followers adjust their movements based on the leader's position and distance to maintain alignment, thus ensuring the overall movement of the multi-robot system. Alternatively, cameras can be installed, using visual algorithms to track the leader's size. This method is suitable for centralized multi-robot architectures but is highly dependent on the leader, exhibiting greater error as followers move further back in the formation, resulting in poor robustness in practical applications. Summary of the Invention

[0004] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a method and apparatus for locating hazardous chemical leak sources in a multi-robot inspection system. The purpose is to control the inspection robots to conduct collaborative inspections in an obstacle-avoidance formation based on autonomous navigation. Each inspection robot includes a lead robot that guides the subsequent inspection robots, while the remaining robots follow the robots in front of them. When any inspection robot detects a hazardous chemical leak source, it issues an audible alarm and sends abnormal gas data to the lead robot via a network. This allows the positioning robot to detach from the formation and search for the accurate location of the leak source, thereby solving the technical problem of poor robustness and inaccurate positioning in existing multi-robot collaborative inspection systems.

[0005] To achieve the above objectives, according to one aspect of the present invention, a method for locating hazardous chemical leak sources in a multi-robot inspection system is provided. The multi-robot inspection system includes a mapping robot, an inspection robot, an acoustic detection robot, and a positioning robot. The method includes:

[0006] S1: Dispatch the mapping robot to draw a working environment map of the multi-robot inspection system; based on the working environment map, disperse and deploy the inspection robots in the initial state to various spaces in the working environment;

[0007] S2: Control each of the inspection robots to conduct collaborative inspections in an obstacle avoidance formation based on autonomous navigation; the inspection robots include a lead robot, which leads the inspection robots behind it, and the other inspection robots follow the inspection robots in front of them.

[0008] S3: When any of the inspection robots detects a hazardous chemical leak source, it issues an audible alarm and sends the abnormal gas data to the navigation robot through the network, so that the navigation robot and the acoustic detection robot can make a preliminary estimate of the location of the hazardous chemical leak source based on the network information and the audible alarm, and issue a search command to the positioning robot.

[0009] S4: Control the positioning robot to search out of the group and execute a biomimetic gas and acoustic fusion search algorithm. Perform preliminary positioning based on the sound alarm detected on site, and then complete the precise positioning of the hazardous chemical leak source based on gas anomaly data and on-site detection.

[0010] In one embodiment, the collaborative inspection in S2 according to the obstacle avoidance formation method based on autonomous navigation includes: autonomous obstacle avoidance and formation formation.

[0011] In one embodiment, the autonomous obstacle avoidance includes:

[0012] Set a sampling range constant sample_view to increase the sampling range, and maintain a minimum radar detection distance min_distance within the sample_view range to expand the detection range. Small obstacles can be detected in advance to prevent getting stuck.

[0013] In one embodiment, the autonomous obstacle avoidance further includes:

[0014] If the inspection robot still has room to move even when it is stuck, record the min_distance for N scanning cycles and store it in the obstacle list, and calculate the variance of min_distance according to the following formula.

[0015]

[0016] Estimate a variance threshold var_thres based on the working environment of the trapped inspection robot. When var_obstacle < var_thres, adjust in the reverse direction of the moving direction and turn simultaneously to break the deadlock.

[0017] In one embodiment, the formation according to the formation includes:

[0018] When the inspection robot follows in a snake formation, the formation deviation is:

[0019]

[0020] The snake formation involves the head machine A, the follower B, and the virtual machine V corresponding to the follower B; where X err is the deviation of the virtual machine V relative to the follower B in the x direction of the coordinate system, and Y err is the deviation of the virtual machine V relative to the follower B in the y direction of the coordinate system; △X is the formation gap between several snake formation robots; θ B is the angle between the line connecting the two points of the head machine A and the follower B and the x-axis.

[0021] In one embodiment, the formation according to the formation includes:

[0022] When the inspection robot forms a triangle formation, the formation deviation is:

[0023]

[0024] The triangle formation involves the head machine A, the follower B, the follower C, and the virtual machine V1 corresponding to the follower B, the virtual machine V2 corresponding to the follower C; where X err1 is the deviation of the virtual machine V1 relative to the follower B in the x direction of the coordinate system, and Y err1 is the deviation of the virtual machine V1 relative to the follower B in the y direction of the coordinate system; X err2 is the deviation of the virtual machine V2 relative to the follower C in the x direction of the coordinate system, and Y err2 is the deviation of the virtual machine V2 relative to the follower C in the y direction of the coordinate system; △X is the formation gap between several triangle formation robots in the x direction, △Y is the formation gap between several triangle formation robots in the y direction; θ B is the angle between the line connecting the two points of the head machine A and the follower B and the x-axis, θ C is the angle between the line connecting the two points of the head machine A and the follower C and the x-axis.

[0025] According to another aspect of the present invention, a hazardous chemical leak source locating device for a multi-robot inspection system is provided, wherein the multi-robot inspection system includes: a mapping robot, an inspection robot, an acoustic detection robot, and a locating robot; the device includes:

[0026] An initialization module is used to dispatch the mapping robot to draw a working environment map of the multi-robot inspection system; based on the working environment map, the inspection robots in the initial state are dispersed and deployed to various spaces in the working environment;

[0027] The inspection module is used to control each of the inspection robots to conduct collaborative inspections in an obstacle avoidance formation based on autonomous navigation. The inspection robots include a lead robot, which acts as the leader to lead the inspection robots behind it, and the other inspection robots follow the inspection robot in front of them.

[0028] The detection module is used to issue an audible alarm when any of the inspection robots detects a hazardous chemical leak source, and to send the abnormal gas data to the navigation robot through the network, so that the navigation robot and the acoustic detection robot can make a preliminary estimate of the location of the hazardous chemical leak source based on the network information and the audible alarm, and issue a search command to the positioning robot.

[0029] The positioning module is used to control the positioning robot to search independently and execute a biomimetic gas and acoustic fusion search algorithm. It performs preliminary positioning based on sound alarms detected on-site, and then completes the precise positioning of the hazardous chemical leak source based on gas anomaly data and on-site detection.

[0030] According to another aspect of the present invention, a multi-robot inspection system is provided, including a mapping robot, an inspection robot, a sound wave detection robot, a positioning robot, a memory, and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0031] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0032] (1) The method for locating hazardous chemical leak sources in the multi-robot inspection system provided in this application controls the inspection robots to conduct collaborative inspections in an obstacle avoidance formation based on autonomous navigation. Each inspection robot includes a lead robot, which acts as the leader, guiding the subsequent inspection robots, while the remaining inspection robots follow the robots in front of them. When any inspection robot detects a hazardous chemical leak source, it issues an audible alarm and sends abnormal gas data to the lead robot via a network, thereby controlling the locating robot to detach from the formation and search for the precise location of the hazardous chemical leak source. All formations are based on the pairwise following behavior of the robots. While maintaining formation, the dependence of the remaining inspection robots on the lead robot within the formation is significantly reduced, improving the robustness of collaborative inspections and ultimately ensuring accurate location. The robots work in groups, with mapping robots, inspection robots, acoustic detection robots, and localization robots cooperating with each other while completing their own tasks. Based on their roles, the robots can be divided into teams and assigned to different areas of the environment. They can share information and complete tasks according to their individual capabilities, thereby significantly improving work efficiency through faster response times and collaborative processing. Furthermore, while the dispersed and collaborative inspection by multiple robots enhances inspection efficiency, it also increases the coverage area of ​​the inspection zone.

[0033] (2) This solution adopts the approach of increasing the radar data observation range of the inspection robot, which can greatly avoid the situation where the robot gets stuck when it encounters small obstacles such as table legs that are not in the robot's forward radar field of view.

[0034] (3) This solution uses the method of maintaining the distance variance of obstacles to solve the deadlock. It can greatly avoid the situation where the algorithm instructs the inspection robot to retreat after encountering an obstacle in front of it, but there is also an obstacle behind it, causing the robot's movement to get stuck.

[0035] (4) In this scheme, the robot responsible for gas acoustic fusion search in a multi-robot cluster searches separately. When the distance is far, a path planning algorithm is executed, and the path is replanned based on the acoustic sensor data after each approach. After determining that it has approached the gas source, a biomimetic gas and acoustic fusion search algorithm is executed to finally locate the leak source. Attached Figure Description

[0036] Figure 1 This is a physical image of the inspection robot developed in this invention.

[0037] Figure 2 This is a flowchart of the multi-robot inspection and positioning process.

[0038] Figure 3This is a flowchart of the autonomous navigation module in multi-machine collaboration.

[0039] Figure 4a This is a flowchart of a formation obstacle avoidance algorithm based on autonomous navigation in multi-machine collaboration.

[0040] Figure 4b This is a schematic diagram of two aircraft following each other in a serpentine formation based on autonomous navigation in multi-aircraft collaboration.

[0041] Figure 5 This is a flowchart of a hazardous chemical leak source search process based on sensor information fusion. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0043] Figure 1 The image shown is a physical representation of the collaborative inspection robot developed in this invention. The multi-robot inspection system operates in groups, with each robot performing its own task while cooperating with others. Each robot can be equipped with different algorithms or sensors to acquire different functions. Based on their job roles, robots can be divided into teams and assigned to different areas of the environment. They can share information and complete tasks according to their individual capabilities, thereby significantly improving work efficiency through faster response times and collaborative processing. While improving inspection efficiency through distributed and collaborative inspection, the system also increases the coverage area of ​​the inspection zone. Furthermore, when a single robot malfunctions, other robots can take over its tasks, allowing the robot group to continue operating stably, demonstrating good robustness, fault tolerance, and stability.

[0044] Figure 2 This is a flowchart illustrating the workflow of a hazardous chemical leak source localization method using a multi-robot inspection system. Specifically, a hazardous chemical leak source localization method based on an autonomous navigation multi-robot inspection system includes the following steps:

[0045] S1: Dispatch the mapping robot to draw a working environment map of the multi-robot inspection system; based on the working environment map, disperse and deploy the inspection robots in the initial state to various spaces in the working environment;

[0046] S2: Control each of the inspection robots to conduct collaborative inspections in an obstacle avoidance formation based on autonomous navigation; the inspection robots include a lead robot, which leads the inspection robots behind it, and the other inspection robots follow the inspection robots in front of them.

[0047] S3: When any of the inspection robots detects a hazardous chemical leak source, it issues an audible alarm and sends the abnormal gas data to the navigation robot through the network, so that the navigation robot and the acoustic detection robot can make a preliminary estimate of the location of the hazardous chemical leak source based on the network information and the audible alarm, and issue a search command to the positioning robot.

[0048] S4: Control the positioning robot to search out of the group and execute a biomimetic gas and acoustic fusion search algorithm. Perform preliminary positioning based on the sound alarm detected on site, and then complete the precise positioning of the hazardous chemical leak source based on gas anomaly data and on-site detection.

[0049] S1: Deploy a robot with mapping capabilities to map the working environment.

[0050] A multi-robot inspection system utilizes LiDAR combined with ROS-based Gmapping to map the working environment, enabling the robots to gain a wider field of view and facilitating subsequent algorithm implementation. During mapping, the robots move slowly through space while ensuring unobstructed radar coverage, allowing real-time observation of the mapping results via the RVIZ program. Once full coverage of the working environment is achieved, a rasterized map is generated and saved, beneficial for robot inspection testing within this map. Subsequently, the map data is shared across the entire system.

[0051] S2: In the initial state, the multi-robot inspection system is deployed in various spaces of the environment. Each robot performs collaborative inspections using an obstacle avoidance formation based on autonomous navigation.

[0052] In the multi-machine collaborative inspection section, this invention designs a formation obstacle avoidance algorithm based on autonomous navigation. This algorithm mainly involves two parts: autonomous obstacle avoidance and formation by formation. The obstacle avoidance capability of the formation algorithm is optimized by equipping the formation with an autonomous navigation obstacle avoidance module.

[0053] First, let's discuss the autonomous navigation module: When using radar algorithms, robots encounter various problems, the two most significant being: 1. The robot may get stuck when it encounters small obstacles, such as table legs, that are not in its forward radar field of view. 2. The algorithm may instruct the robot to retreat after encountering an obstacle in front, but if there is also an obstacle behind it, the robot's movement may become stuck. To address these two issues, improvements to the autonomous navigation module are needed:

[0054] For the first case, a solution of increasing the radar data observation range is adopted. A sampling range constant sample_view is set to increase the sampling range, and a minimum detection distance min_distance of the radar within the sample_view range is maintained. This can expand the detection range, and small-sized obstacles can be detected in advance to prevent getting into a deadlock.

[0055] For the second case, if the robot still has movement space when it is stuck, this design uses the method of maintaining the variance of the obstacle distance to rescue the deadlock. The specific operation is as follows: The min_distance of N scanning cycles is recorded and stored in the obstacle list, and the variance var_obstacle of min_distance is calculated according to the following formula:

[0056]

[0057] Then, a variance threshold var_thres is estimated according to the robot working environment. When var_obstacle < var_thres, it means that the distance between the robot and the obstacle within N cycles, that is, the position of the robot, may not change significantly, and it may belong to the situation of getting into a deadlock in the above second algorithm. Then, it actively adjusts in the reverse direction of the movement direction and turns at the same time to break free from the deadlock. It should be noted that the condition that var_obstacle is greater than a very small number should also be added at the same time to avoid unnecessary "breaking free from the deadlock" operations when var_obstacle is 0 when there are no obstacles around the robot.

[0058] In summary, the operation process of the adjusted autonomous navigation algorithm is as Figure 3 shown.

[0059] Next, the inspection formation design is introduced:

[0060] First, a basic multi-robot formation algorithm is introduced: the multi-robot formation algorithm based on the consistency Leader-Follower. All formations are based on the following behavior between two robots. This algorithm assumes that there are virtual robots following constantly behind the leader robot, and the actual follower robots move continuously towards the virtual robots through motion control to achieve the following effect of maintaining the formation.

[0061] Assume the leader robot is A, the follower robot is B, and the virtual robot is V. Then the position of the follower robot B relative to A in the world coordinates is:

[0062]

[0063] Similarly, the position of the virtual machine V relative to A in the world coordinates is:

[0064]

[0065] Based on the positional deviation of virtual machine V relative to B:

[0066] The method for calculating the deviation within the coordinate system of aircraft B can be described as follows:

[0067]

[0068] This allows us to determine the positional relationship of the deviation centered at B. Then, based on the motion model of A, let the forward velocity of A be v. A The clockwise speed is ω A Solving the above equations simultaneously and taking their derivatives, we can obtain the following:

[0069] Based on the above formula, the motion controller is configured for compensation tracking, allowing the follower B to continuously approach the virtual machine V:

[0070] Where K1, K2, and K3 are adjustable constant coefficients, and in the actual execution of the algorithm, ω A The assistance provided for following is minimal and can be omitted. This leads to the robot following process in the navigation and following algorithm.

[0071] The aforementioned algorithm suffers from poor system robustness because all robots follow the same lead robot, albeit with relative positional deviations. To address this issue, this invention proposes a distributed following solution: a multi-robot system has an overall leader, with subsequent robots following the leader in a specific formation while also being followed by the robots behind. This allows the entire system to move in a more stable cluster state, improving formation robustness. The following analysis uses serpentine and triangular formations as examples to illustrate formation design. The workflow of the multi-robot system formation obstacle avoidance algorithm is as follows: Figure 4a As shown.

[0072] like Figure 4b As shown, to complete the serpentine following pattern, the deviation X in the X direction is... err To add a ΔX as the deviation, where ΔX is the spacing between the serpentine formation robots, we can obtain the serpentine formation deviation formula:

[0073]

[0074] Similarly, when designing a triangular formation, the relative coordinate position deviation of the follower vertices relative to the leader vertices must be considered. Therefore, if an isosceles triangular formation is designed, the formulas for the left and right follower points B and C are:

[0075] After designing the formation, referencing the idea of ​​behavior-based formation algorithms, multiple robots can make obstacle avoidance responses while forming the formation. In the formation, each robot keeps following while adding an obstacle avoidance module.

[0076] S3: The robot detects a hazardous chemical leak and issues an alarm. The alarm and abnormal data are then transmitted to the lead robot via a network-based system.

[0077] In a multi-robot inspection system, information transmission between robots is crucial for collaborative operation. This system employs a distributed architecture with a control center. During collaborative inspections, the main PC at the control center communicates with each robot in a distributed manner. This invention utilizes the ROS architecture to achieve communication between the PC and the robots, and among the robots themselves.

[0078] During collaborative inspections, the robot's onboard gas and acoustic sensors continuously sample the environment. If any data anomalies are detected, an alarm and the abnormal data are sent to the lead robot, which then processes the information.

[0079] S4: In a multi-robot cluster, the robot responsible for gas-acoustic fusion search searches out of the group, executes a biomimetic gas-acoustic fusion search algorithm, and completes the location of the hazardous chemical leak source.

[0080] Based on the abnormal data, the approximate location of the gas leak source is first calculated collaboratively by multiple robots. After obtaining the location, the robot responsible for gas and acoustic fusion search in the multi-robot cluster decouples from the group to search separately. When the distance is relatively far, a path planning algorithm is executed, and the path is replanned based on the acoustic sensor data after each approach. Once it is determined that the gas source has been approached, a biomimetic gas and acoustic fusion search algorithm is executed to finally locate the leak source. Specifically, the flowchart of the hazardous chemical leak source search based on sensor information fusion is as follows: Figure 5 As shown.

[0081] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for locating hazardous chemical leak sources in a multi-robot inspection system, characterized in that, The multi-robot inspection system includes a mapping robot, an inspection robot, a sound wave detection robot, and a positioning robot; the method includes: S1: Dispatch the mapping robot to draw a working environment map of the multi-robot inspection system; based on the working environment map, disperse and deploy the inspection robots in their initial states to each space of the working environment; S2: Control each of the inspection robots to perform collaborative inspections in an obstacle avoidance formation manner based on autonomous navigation, including autonomous obstacle avoidance; the inspection robot includes a leading robot, which acts as a leader to lead the inspection robots behind it, and the remaining inspection robots follow the inspection robots in front of them; The autonomous obstacle avoidance includes: Set a sampling range constant sample_view, increase the sampling range, and maintain a minimum radar detection distance min_distance within the sample_view range to expand the detection range, so that small obstacles can be detected in advance and prevent getting into a deadlock; If the inspection robot still has room for movement when it is stuck, record the min_distance of N scanning cycles and store it in the obstacle list, and calculate the variance var_obstacle of min_distance according to the following formula: ； Estimate a variance threshold var_thres based on the working environment of the trapped inspection robot. When var_obstacle < var_thres, adjust in the reverse direction of the movement direction and turn at the same time to break free from the deadlock; S3: When any one of the inspection robots detects a hazardous chemical leakage source, issue a sound alarm, and send the gas anomaly data to the leading robot through networking, so that the leading robot and the sound wave detection robot can preliminarily estimate the location of the hazardous chemical leakage source based on the networking information and the sound alarm, and send a search instruction to the positioning robot; S4: Control the positioning robot to break away from the team and search, and execute a bionic gas and acoustic fusion search algorithm. Perform preliminary positioning based on the sound alarm detected on the spot, and then complete the precise positioning of the hazardous chemical leakage source based on the gas anomaly data and the on-site detection.

2. The method for locating hazardous chemical leak sources in a multi-robot inspection system as described in claim 1, characterized in that, The control of each of the inspection robots to perform collaborative inspections in an obstacle avoidance formation manner based on autonomous navigation includes forming a formation according to the formation; the formation according to the formation includes: when the inspection robot follows in a snake formation, the formation deviation is: ； The serpentine formation involves a lead machine A, follower machines B, and the virtual machine V corresponding to follower machine B; where X err Let Y be the deviation of virtual machine V relative to follower B in the x-direction of the coordinate system. err ΔX represents the deviation of virtual machine V relative to follower B in the y-axis of the coordinate system; ΔX represents the difference in formation between several snake-like formation robots. B Let x be the angle between the line connecting the two points of the head machine A and the follower machine B and the x-axis.

3. The method for locating hazardous chemical leak sources in a multi-robot inspection system as described in claim 2, characterized in that, The formation according to the formation includes: When the inspection robot forms a triangle formation, the formation deviation is: ； The triangle formation involves a lead machine A, follower machines B and C, as well as virtual machine V1 corresponding to follower machine B and virtual machine V2 corresponding to follower machine C; where X err1 Let Y be the deviation of virtual machine V1 relative to follower B in the x-direction of the coordinate system. err1 X represents the deviation of virtual machine V1 relative to follower B in the y-direction of the coordinate system; err2 Let Y be the deviation of virtual machine V2 relative to the follower machine C in the x-direction of the coordinate system. err2 ΔX represents the deviation of virtual machine V2 relative to follower C in the y-direction of the coordinate system; ΔX represents the difference in formation between several triangular formation robots in the x-direction, and ΔY represents the difference in formation between several triangular formation robots in the y-direction. B Let the angle between the line connecting the two points of the head machine A and the follower machine B and the x-axis be... C Let x be the angle between the line connecting the head machine A and the follower machine C and the x-axis.

4. A hazardous chemical leak source locating device for a multi-robot inspection system, characterized in that, The multi-robot inspection system includes: a mapping robot, an inspection robot, a sound wave detection robot, and a positioning robot; for performing the method for locating a hazardous chemical leakage source of the multi-robot inspection system according to any one of claims 1-3, the device includes: An initialization module, configured to dispatch the mapping robot to draw a working environment map of the multi-robot inspection system; based on the working environment map, disperse and deploy the inspection robots in their initial states to each space of the working environment; The inspection module is used to control each of the inspection robots to conduct collaborative inspections in an obstacle avoidance formation based on autonomous navigation. The inspection robots include a lead robot, which acts as the leader to lead the inspection robots behind it, and the other inspection robots follow the inspection robot in front of them. The detection module is used to issue an audible alarm when any of the inspection robots detects a hazardous chemical leak source, and to send the abnormal gas data to the navigation robot through the network, so that the navigation robot and the acoustic detection robot can make a preliminary estimate of the location of the hazardous chemical leak source based on the network information and the audible alarm, and issue a search command to the positioning robot. The positioning module is used to control the positioning robot to search independently and execute a biomimetic gas and acoustic fusion search algorithm. It performs preliminary positioning based on sound alarms detected on-site, and then completes the precise positioning of the hazardous chemical leak source based on gas anomaly data and on-site detection.

5. A multi-robot inspection system, comprising a mapping robot, an inspection robot, a sound wave detection robot, a positioning robot, a memory, and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 3.

Images