An automatic patrol fire-fighting robot based on close detection strategy and a detection method

By using an automated inspection and firefighting robot based on a proximity detection strategy, the robot utilizes binocular vision to measure the azimuth angle of the fire source and plan a reference path. This solves the problem of inaccurate fire source detection in existing technologies and enables accurate path planning and rapid response to fire sources within different confidence intervals.

CN116407797BActive Publication Date: 2026-03-20BEIJING MUNICIPAL ZHONGYAN CONSTR MASCH MFG CO LTD
View PDF 35 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-01-29
Publication Date
2026-03-20

AI Technical Summary

Technical Problem

Existing automated inspection firefighting robots have difficulty accurately detecting the location of fire sources at a distance, resulting in the loss of fire source signals and slow detection speed, making it impossible to respond to fire situations in a timely manner.

Method used

An automated inspection firefighting robot based on a proximity detection strategy measures the azimuth angle of the fire source using binocular vision, plans a reference path, and approaches the fire source for detection. It also optimizes the path planning by combining confidence interval labels to ensure that the fire source signal is not lost and improves detection efficiency.

Benefits of technology

It achieves accurate path planning within different confidence intervals, avoids loss of fire source signals, ensures that the robot can track the fire source in a timely manner, and improves the accuracy and response speed of fire source detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116407797B_ABST
    Figure CN116407797B_ABST
Patent Text Reader

Abstract

The application discloses an automatic patrol fire-fighting robot based on a close detection strategy and a detection method, and comprises the following steps: step one, building an automatic patrol fire-fighting robot with an unmanned driving function; step two, integrating an electric control fire extinguishing device; and step three, designing a fire source close detection strategy and deploying the strategy on the upper part of the automatic patrol fire-fighting robot. Path planning is carried out in different confidence intervals in stages, so that the loss of fire source signals caused by the inaccurate measurement of the positions of fire sources far away can be reduced. Secondly, when the close strategy is executed, the target point of the automatic patrol strategy is changed to the position of the fire source, and the heading at the target point should be consistent with the current heading of the robot, so that the robot can be prevented from moving away from the fire source after the loss of the fire source signals, and the robot can be prevented from constantly planning a reference path after receiving the fire source signals. Finally, the above strategy considers the situation that a fire occurs at a high position, and the loss of the fire source signals caused by the high position of the fire source can be avoided.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of fire-fighting robots, and particularly relates to an automatic patrol fire-fighting robot based on a close detection strategy and a detection method. BACKGROUND

[0002] Firefighting robots are an important device that can reduce the risk of fire. Early firefighting robots are usually visual range remote control or remote control semi-automatic robots (CN103706071A; CN111481862B; CN205287342U; CN206762078U), which need to be operated by workers and cannot achieve automatic inspection. With the development of automation technology, firefighting robots with automatic inspection function have gradually appeared. Automatic inspection firefighting robots contain two key technologies: one is unmanned driving technology, and the other is fire detection technology. Due to the rapid development of unmanned driving technology of unmanned vehicles and other mobile equipment, unmanned driving technology is no longer a technical bottleneck for automatic inspection firefighting robots. However, there are still many problems to be solved in fire detection technology. In the existing firefighting robots with automatic inspection function, some firefighting robots do not disclose the fire detection scheme (CN105096507B; CN114129941A; CN111481865A; CN202113524U; CN206792856U; CN209900500U), and some firefighting robots use flame sensors and range finders to detect the position of the fire (Li B, Chen D Y, Ma X Y. Design and implementation of intelligent fire extinguishing robot system [J]. Machinery and electronics, 2010 (1): 57-61; Li X Y, Chen D Y, Ma X Y. Design and implementation of intelligent fire extinguishing robot [J]. Electronic design engineering, 2010, 18 (3): 51-54.CN107115613A; CN110812742A; CN112911157A; CN114792339A; CN202654580U; CN212880710U; CN214277903U; CN216246838U), wherein the flame sensor can be a camera, an infrared thermal imager, an infrared temperature measuring instrument, etc., and the range finder can be a laser range finder, an infrared range finder, an ultrasonic range finder, etc. Since the open flame is a kind of plasma, it is usually difficult for the range finder to measure the accurate distance of the plasma, and therefore it is also difficult for this detection scheme to obtain the accurate position of the fire source. Some firefighting robots use a rotating infrared thermal imager to detect the fire source (CN111599129A; CN111617414A; CN113521616A; CN111617415A; CN112911157A; CN205182046U; CN214025693U; CN216803447U), and the disadvantage of this detection scheme is that there are more moving parts and higher requirements for rotation control accuracy, and it is prone to failure under harsh conditions such as firefighting operations. Some firefighting robots increase binocular vision to measure the distance of the fire source relative to the robot on the basis of this detection scheme (CN111443711A; CN111599129A), and this scheme also has the problems of more parts and higher requirements for rotation control accuracy. Some firefighting robots directly use binocular vision to measure the position of the fire source (LiG, LuG, YanY. Fire detection using stereoscopic imaging and image processing techniques [C] / / 2014 IEEE International Conference on Imaging Systems and Techniques (IST) Proceedings.Santorini, Greece, 2014: 28-32; CN105261029A; CN110201333A; CN110180114A; CN111408089A; CN113413564A; CN114200471A; CN114432635A; CN212439798U), wherein the binocular camera can be all visible light cameras, can be all infrared thermographs, or can be one visible light camera and the other infrared thermograph, the main disadvantage of this detection scheme is that the two cameras are relatively close, making it difficult to measure the accurate fire source position when the fire source is far away, to solve this problem, a triangulation method appears, this detection scheme is also based on the principle of binocular vision, specifically, the robot first measures the azimuth angle of the fire source after discovering the fire source, then the robot drives a distance and measures the azimuth angle of the fire source again, and calculates the position of the fire source according to the trigonometric function (CN105954718A; CN113819881A), this detection scheme can measure the accurate position of the fire source far away, but the detection speed is slow, and the robot cannot respond to the fire in time. In summary, the fire source detection scheme using binocular vision has the advantages of few moving parts and fast detection speed, but cannot measure the accurate position of the fire source far away, which may cause the automatic patrol fire-fighting robot to lose the fire source signal when tracking the fire source. SUMMARY

[0003] The technical scheme adopted by the present application is as follows: An automatic patrol fire-fighting robot based on a close-in detection strategy, comprising:

[0004] A device table;

[0005] A running structure arranged at both sides of the outer wall of the device table, wherein: the running structure is provided with two groups, and each group of the running structure comprises a track, a buffer plate, a buffer rod, a supporting wheel, a buffer spring, a buffer wheel, a guide wheel, a driving wheel and a chain wheel, the buffer plate is fixedly arranged at one side of the outer wall of the device table through a support, the buffer rod is rotatably arranged at the edge of one side of the outer wall of the buffer plate, the supporting wheel is rotatably embedded at one end of the buffer rod through a rotating shaft, one end of the buffer spring is rotatably inserted into the outer wall of the buffer plate, the other end of the buffer spring is rotatably inserted into the outer wall of the buffer rod, the buffer wheel is rotatably inserted into one side of the outer wall of the buffer plate, the guide wheel is rotatably inserted into one side of the outer wall of the buffer plate, the driving wheel is rotatably inserted into one side of the outer wall of the buffer plate, and the chain wheel is rotatably inserted into one side of the outer wall of the buffer plate, and the track and the driving wheel are in driving connection with each other.

[0006] Further, the inner wall of the track is provided with an induction tooth, and the induction tooth, the supporting wheel, the buffer wheel, the guide wheel and the chain wheel are matched with each other.

[0007] Further, the outer wall top side of the equipment table is fixedly provided with a mounting table, the outer wall top of the equipment table is fixedly provided with a camera, and the outer wall top of the equipment table is fixedly provided with an antenna.

[0008] A detection method of an automatic patrol fire-fighting robot based on a close-in detection strategy, comprising the following steps:

[0009] S1, building an automatic patrol fire-fighting robot with an unmanned driving function.

[0010] S2, integrating an electric control fire extinguishing device.

[0011] S3, designing a fire source close-in detection strategy and deploying it on the automatic patrol fire-fighting robot.

[0012] Further, in step S1, the automatic patrol fire-fighting robot comprises an industrial computer, a laser radar, a GPS and an IMU.

[0013] Further, in step S2, the electric control fire extinguishing device refers to a fire extinguishing device activated by electric signals such as current and voltage.

[0014] Further, in step S3,

[0015] S301, judging whether there is a fire source in the binocular vision field, if yes, reporting a fire alarm and a fire source position (the position may not be accurate, but the confidence will gradually increase as the distance is close) to the command center, and executing S302, if not, executing an automatic patrol strategy, i.e. a strategy of following a preset reference trajectory.

[0016] S302, judging whether the straight-line distance between the fire source and the robot is greater than the operation threshold distance, the operation threshold distance being the effective operation range of the electric control fire extinguishing device, if yes, executing S303, if not, executing fire extinguishing operation, and sending a start instruction to the electric control fire extinguishing device by the controller.

[0017] S303, judging whether the fire source is lower than the elevation threshold of the vision field, the elevation threshold of the vision field being calibrated by the effective operation range of the electric control fire extinguishing device, i.e. when the electric control fire extinguishing device works at the maximum elevation angle and the straight-line distance between the fire source and the robot is within the effective operation range of the electric control fire extinguishing device, the pixel elevation of the center of the fire source in the binocular camera vision field from the center of the vision field, if yes, executing S304, if not, stopping automatic operation, reporting a fire alarm to the command center, and waiting for the command center instruction.

[0018] S304, judging the confidence interval of the fire source distance, which can be divided into four, arranged from low to high.

[0019] Further, the command center instruction, video information and start instruction of the electric control fire extinguishing device are sent by the antenna.

[0020] Further, in step S304, when the distance is greater than 30m, the confidence is the lowest, and the first confidence interval is obtained; when the distance is greater than 20m and less than 30m, the confidence is the second lowest, and the second confidence interval is obtained; when the distance is greater than 10m and less than 20m, the confidence is the second highest, and the third confidence interval is obtained; and when the distance is less than 10m, the confidence is the highest, and the fourth confidence interval is obtained.

[0021] Further, in step S304, when the distance is greater than 30m, the confidence is the lowest, and the first confidence interval is obtained; when the distance is greater than 20m and less than 30m, the confidence is the second lowest, and the second confidence interval is obtained; when the distance is greater than 10m and less than 20m, the confidence is the second highest, and the third confidence interval is obtained; and when the distance is less than 10m, the confidence is the highest, and the fourth confidence interval is obtained.

[0022] In summary, due to the adoption of the above technical solutions, the present application has the following beneficial effects:

[0023] In the present application, firstly, path planning is performed in different confidence intervals in stages, which can reduce the situation of fire signal loss caused by inaccurate measurement of fire source position at a long distance. Secondly, when the approaching strategy is executed, the target point of the automatic patrol strategy is changed to the fire source position, and the heading at the target point should be consistent with the current heading of the robot, which can avoid the robot moving away from the fire source after the fire signal is lost, and is beneficial to the robot detecting the fire signal again during the approaching process. In addition, the confidence interval label is set, which can avoid the robot constantly planning the reference path after receiving the fire signal. Finally, the above strategy considers the situation of fire at a high place, and can avoid the problem of fire signal loss caused by the high position of the fire source. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 It is a perspective view of the track of the present application;

[0025] Figure 2 It is a perspective view of the buffer spring of the present application;

[0026] Figure 3 It is a schematic diagram of the implementation process of the present application;

[0027] Figure 4 Flow chart for the fire source proximity detection strategy of the present application;

[0028] Figure 5 Flow chart for the confidence interval detection strategy of the present application.

[0029] Marked in the figure: 1, equipment table; 2, installation table; 3, camera; 4, antenna; 5, track; 6, buffer plate; 7, buffer rod; 8, support wheel; 9, buffer spring; 10, buffer wheel; 11, guide wheel; 12, drive wheel; 13, chain wheel. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0031] Example one

[0032] Reference Figure 1 - Figure 5 An automatic patrol fire-fighting robot based on a proximity detection strategy, comprising:

[0033] Equipment table 1;

[0034] A running structure is arranged at both sides of the outer wall of the equipment table 1, wherein: the running structure is provided with two groups, each group of running structure includes track 5, buffer plate 6, buffer rod 7, support wheel 8, buffer spring 9, buffer wheel 10, guide wheel 11, drive wheel 12 and chain wheel 13, the buffer plate 6 is fixedly arranged at one side of the outer wall of the equipment table 1 through the support, the buffer rod 7 is rotatably arranged at the edge of one side of the outer wall of the buffer plate 6, the support wheel 8 is rotatably embedded at one end of the buffer rod 7 through the rotating shaft, one end of the buffer spring 9 is rotatably inserted into the outer wall of the buffer plate 6, the other end of the buffer spring 9 is rotatably inserted into the outer wall of the buffer rod 7, the buffer wheel 10 is rotatably inserted into one side of the outer wall of the buffer plate 6, the guide wheel 11 is rotatably inserted into one side of the outer wall of the buffer plate 6, the drive wheel 12 is rotatably inserted into one side of the outer wall of the buffer plate 6, the chain wheel 13 is rotatably inserted into one side of the outer wall of the buffer plate 6, the track 5 and the drive wheel 12 are drivingly connected with each other, through the running structure, stable driving running can be realized, in the subsequent use process, the buffer spring 9 can offset the shaking of the buffer wheel 10, the guide wheel 11 and the support wheel 8, the buffering effect is achieved, and the automatic patrol fire-fighting robot can quickly reach the fire scene.

[0035] Reference Figure 1 - Figure 5The inner wall of the track 5 is provided with an induction tooth, and the induction tooth, the supporting wheel 8, the buffer wheel 10, the guide wheel 11 and the chain sprocket 13 are matched with each other, the outer wall top side of the equipment table 1 is fixedly provided with the mounting table 2, the mounting table can conveniently install the electric control fire extinguishing device, the outer wall top of the equipment table 1 is fixedly provided with the camera 3, stable image information transmission can be realized through the camera, and the outer wall top of the equipment table 1 is fixedly provided with the antenna 4.

[0036] With reference to Figure 1 - Figure 5A detection method for an automatic inspection fire-fighting robot based on a proximity detection strategy includes the following steps: S1, constructing an automatic inspection fire-fighting robot with unmanned driving capabilities, the automatic inspection fire-fighting robot including an industrial control computer, lidar, GPS and IMU; S2, integrating an electrically controlled fire extinguishing device, which refers to a fire extinguishing device activated by electrical signals such as current and voltage; S3, designing a fire source proximity detection strategy and deploying it on the automatic inspection fire-fighting robot, the provided fire source proximity detection strategy is based on the principle of binocular vision (LiG, LuG, YanY. Fire detection using stereoscopic imaging and image processing techniques [C] / / 2014 IEEE International Conference on Imaging Systems and Techniques (IST) Proceedings. Sant Based on Orini, Greece, 2014: 28-32, the principle is public knowledge and will not be elaborated here. The specific implementation of binocular vision can adopt a deep learning-based method, which is also public knowledge and will not be elaborated here. Electrically controlled fire extinguishing devices refer to fire extinguishing devices activated by electrical signals such as current and voltage. Fire extinguishing devices should have characteristics such as being portable and small in size.The electric control fire extinguishing device is also the basis for implementing the patent S301, judging whether there is a fire source in the binocular vision field of view, if yes, reporting the fire alarm and the fire source position to the command center. The position may not be accurate, but as the distance approaches, the confidence will gradually increase, and the command center instructions, video information and the starting instructions of the electric control fire extinguishing device are sent out by the antenna 4 S302, if no, executing the automatic patrol strategy, that is, the strategy of following the preset reference track, S302, judging whether the straight line distance between the fire source and the robot is greater than the operation threshold distance, that is, the effective operation range of the electric control fire extinguishing device, if yes, executing S303, if no, executing the fire extinguishing operation, sending the starting instructions to the electric control fire extinguishing device by the controller, S303, judging whether the fire source is lower than the field of view elevation threshold, the field of view elevation threshold is calibrated by the effective operation range of the electric control fire extinguishing device, that is, when the electric control fire extinguishing device works at the maximum elevation angle and the straight line distance between the fire source and the robot is within the effective operation range of the electric control fire extinguishing device, the pixel elevation of the center of the fire source in the binocular camera field of view from the center of the field of view, if yes, executing S304, if no, stopping the automatic operation, reporting the fire alarm to the command center, waiting for the command center instructions, S304, judging the confidence interval of the distance of the fire source, the confidence interval can be divided into four, arranged from low to high, when the distance is greater than 30 m, the confidence is the lowest, which is the first confidence interval, when the distance is greater than 20 m and less than 30 m, the confidence is the second lowest, which is the second confidence interval, when the distance is greater than 10 m and less than 20 m, the confidence is the second highest, which is the third confidence interval, when the distance is less than 10 m, the confidence is the highest, which is the fourth confidence interval, when the distance of the fire source is in the confidence interval, first, the inaccurate fire source position measured by the binocular vision is taken as the target point to plan the reference path, in particular, the heading at the target point should be consistent with the current heading of the robot, then the robot follows the reference path, in addition, when the robot follows the reference path, the fire source in the field of view may be lost, therefore, after the robot enters the strategy, the target point of the automatic patrol strategy needs to be modified to the fire source position, the heading at the target point should be consistent with the current heading of the robot, in addition, considering that the path planning needs a certain time, if the binocular vision signal executes the path planning once after meeting the confidence interval each time, the driving efficiency of the robot will be greatly reduced, therefore, after the robot enters the confidence interval, a confidence interval label needs to be set, the label is a global variable, the initial value is 0, after entering the confidence interval, first, the judgment is made, if the value of the label is 0, the value of the label is set to 1, then the path planning is executed, and the path tracking program is called, if the label is 1, the existing reference path is called, and the path tracking program is called.

[0037] The above merely describes the preferred embodiments of the present application and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A detection method for an automatic inspection firefighting robot based on a proximity detection strategy, characterized in that, Includes the following steps: S1. Build an automated inspection and firefighting robot with unmanned driving capabilities; S2, Integrated electronically controlled fire extinguishing device; S3. Design a fire source proximity detection strategy and deploy it on an automated inspection firefighting robot; In step S3, S301. Determine if there is a fire source within the binocular visual field. If so, report the fire alarm and the location of the fire source to the command center and execute S302. If not, execute the automatic patrol strategy, i.e., follow the preset reference trajectory. S302. Determine whether the straight-line distance between the fire source and the robot is greater than the working threshold distance. The working threshold distance is the effective working range of the electronically controlled fire extinguishing device. If yes, execute S303. If no, execute the fire extinguishing operation and send a start command to the electronically controlled fire extinguishing device from the controller. S303. Determine whether the fire source is below the field of view elevation threshold. The field of view elevation threshold is determined by the effective working range of the electronically controlled fire extinguishing device. That is, when the electronically controlled fire extinguishing device is working at its maximum elevation angle and the straight-line distance between the fire source and the robot is within the effective working range of the electronically controlled fire extinguishing device, the pixel elevation of the fire source center in the field of view of the binocular camera is the distance from the center of the field of view. If yes, execute S304. If no, stop automatic operation, report the fire alarm to the command center, and wait for instructions from the command center. S304. Determine the confidence interval for the distance to the fire source. The confidence interval can be divided into 4 intervals, arranged from low to high confidence. In step S304, when the distance to the fire source is within the confidence interval, the target point of the automatic patrol strategy is modified to the location of the fire source. The heading at the target point should be consistent with the current heading of the robot. The reference path is planned using the inaccurate fire source location measured by binocular vision as the target point. After the robot enters the confidence interval, a confidence interval label is set. This label is a global variable with an initial value of 0. After entering the confidence interval, a judgment is first made. If the label value is 0, the label value is set to 1. Then, path planning is performed and the path tracking program is called. If the label value is 1, the existing reference path is called and the path tracking program is called. In step S304, the confidence level is lowest when the distance is greater than 30m, which is the first confidence interval; the confidence level is second lowest when the distance is greater than 20m and less than 30m, which is the second confidence interval; the confidence level is second highest when the distance is greater than 10m and less than 20m, which is the third confidence interval; and the confidence level is highest when the distance is less than 10m, which is the fourth confidence interval.

2. The detection method for an automatic inspection firefighting robot based on a proximity detection strategy as described in claim 1, characterized in that, In step S1, the automatic inspection firefighting robot includes an industrial control computer, a lidar, a GPS, and an IMU.

3. The detection method for an automatic inspection firefighting robot based on a proximity detection strategy as described in claim 2, characterized in that, In step S2, the electrically controlled fire extinguishing device refers to a fire extinguishing device activated by current and voltage electrical signals.

4. The detection method for an automatic inspection firefighting robot based on a proximity detection strategy as described in claim 3, characterized in that, Command center instructions, video information, and activation commands for the electronically controlled fire extinguishing device are transmitted via the antenna.

Citation Information

Patent Citations

  • Fire-fighting robot

    CN103706071A

  • Fire-fighting robot video perception patrol alarm device and patrol alarm method

    CN105096507B

  • Fire scene fire source positioning method and system and fire-fighting robot

    CN105954718A

  • Fire-fighting robot co-location, reconnaissance, fire source recognition and aimed fire extinguishing method

    CN110180114A

  • Full-automatic cooperative reconnaissance and fire extinguishment working method of fire-fighting robot

    CN110201333A