An autonomous operation control method and system of a highway tunnel fire extinguishing quadruped robot

Abstract

The present application relates to the technical field of robot path planning, and proposes an autonomous operation control method and system for a highway tunnel fire extinguishing quadruped robot, comprising: collecting 3D fire scene maps and two-dimensional tunnel maps of the highway tunnel at different collection times; marking target data points when a fire occurs, determining the temperature variation degree of the target data points and the fire source characteristic values of each data point in the 3D fire scene map; determining the fire source target points and the fire area, retaining part of the fire source target points, and determining the fire extinguishing quadruped robot for extinguishing the fire source target points; determining the adjustment step length and the fire extinguishing path of the fire extinguishing quadruped robot, and the fire extinguishing quadruped robot travels according to the fire extinguishing path and reaches the fire extinguishing position; and the fire extinguishing quadruped robot completes the autonomous operation control of fire extinguishing at the fire extinguishing position. The present application can improve the rationality of the path planning of the fire extinguishing quadruped robot in the highway tunnel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robot path planning, in particular to an autonomous operation control method and system of a highway tunnel fire extinguishing quadruped robot. BACKGROUND

[0002] Highway tunnel fires have the characteristics of closed environment, low visibility, large temperature gradient, rapid spread, etc., and traditional fire-fighting equipment is difficult to efficiently reach the core fire area. The fire extinguishing quadruped robot can be deployed in the tunnel or at the entrance in advance, quickly respond to the fire, and achieve accurate positioning and fire extinguishing. Path planning is the core of the autonomous operation of the fire extinguishing quadruped robot, and the RRT fast random expansion tree algorithm can be used to achieve path planning.

[0003] However, the RRT algorithm usually uses a preset fixed step size to achieve path planning. When the step size is too large, the fire extinguishing quadruped robot will frequently collide in the complex areas of the highway tunnel, resulting in a large number of sample points being discarded, the algorithm efficiency being too low, and even the situation that a feasible path cannot be found may occur. When the step size is too small, the fire extinguishing quadruped robot will slowly crawl in the open areas of the highway tunnel, resulting in low exploration efficiency, an increased risk of fire spreading, and possibly leading to problems such as non-smooth exploration path and difficult control execution. SUMMARY

[0004] The present application provides an autonomous operation control method and system of a highway tunnel fire extinguishing quadruped robot to solve the problem that the RRT algorithm uses a fixed step size to achieve path planning and cannot meet the needs of different conditions in the highway tunnel. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present application provides an autonomous operation control method of a highway tunnel fire extinguishing quadruped robot, which comprises the following steps:

[0006] According to the preset sensors in the highway tunnel to be completed for the fire extinguishing task, 3D fire field maps and two-dimensional tunnel maps of the highway tunnel at the current collection time and the adjacent different collection times before are obtained, and the data points in the 3D fire field maps correspond to spatial position coordinates and temperature data;

[0007] According to the 3D fire field map of the highway tunnel, it is determined whether a fire occurs. If a fire occurs, any one data point in the 3D fire field map of the highway tunnel is recorded as a target data point. According to the difference between the temperature data of the target data point at the current collection time and the adjacent collection times before, the temperature variation degree of the target data point at the current collection time is determined. According to the positions of the adjacent data points in the 3D fire field map at the current collection time, the temperature data of each data point, and the temperature variation degree, the fire source characteristic value of each data point in the 3D fire field map is determined.

[0008] According to the difference between the position distribution and the temperature data of all data points in the 3D fire field map at the current collection time, the fire source target point and the fire area are determined, according to the number of data points in the fire area where the fire source target point is located, the temperature change degree and the fire source characteristic value, and the position relationship between the fire extinguishing quadruped robot and the fire source target point, part of the fire source target points are reserved, and the fire extinguishing quadruped robot for extinguishing the fire source target point is determined;

[0009] According to the initial value of the parameter of the RRT algorithm set according to the position of the fire extinguishing quadruped robot, the local node path is obtained, the adjustment step of the fire extinguishing quadruped robot is determined in combination with the position relationship between the fire extinguishing quadruped robot and the fire source target point, and the fire extinguishing path of the fire extinguishing quadruped robot is obtained according to the adjustment step, and the fire extinguishing quadruped robot travels according to the fire extinguishing path and reaches the fire extinguishing position;

[0010] The fire extinguishing quadruped robot completes the autonomous operation control of extinguishing the fire at the fire extinguishing position.

[0011] Further, the method for determining whether a fire occurs according to the 3D fire field map of the highway tunnel comprises the following specific method:

[0012] The mean value of the temperature data of all data points in the 3D fire field map is recorded as the temperature mean value, and when the temperature data of the data points in the 3D fire field map is greater than or equal to twice the temperature mean value, it is determined that a fire occurs in the highway tunnel.

[0013] Further, the determination of the temperature change degree of the target data point at the current collection time is as follows:

[0014] Any one collection time within the previous minutes and the current collection time is recorded as the target collection time, and the absolute value of the difference between the temperature data of the target data point at the target collection time and the temperature data of the target data point at the adjacent collection time before the target collection time is recorded as the first absolute value of the target data point at the target collection time.

[0015] The product of the cumulative sum of the first absolute values of the target data point at all collection times within the previous minutes and the current collection time and the temperature data of the target data point at the current collection time is recorded as the temperature change degree of the target data point at the current collection time; wherein, represents the first preset threshold.

[0016] Further, the specific method for obtaining the fire source characteristic value of each data point in the 3D fire field map is as follows:

[0017] The data points with the closest Euclidean distance to the target data point in the 3D fire field map at the current collection time are recorded as the target data points, and the temperature data of the target data points are recorded as the temperature data of the target data points. The difference between the temperature data of the target data point and the adjacent data point is denoted as the temperature change degree of the target data point, and the direction from the target data point to the adjacent data point is denoted as the extension direction of the adjacent data point. For all adjacent data points in the extension direction of the adjacent data point, the difference between the temperature data of the adjacent adjacent data point and the temperature change degree of the previous adjacent data point is denoted as the first product of the previous adjacent data point. Wherein, represents a second preset threshold value;

[0018] The average of the first products of all adjacent data points in the extension direction of the adjacent data point is denoted as the first average of the adjacent data point.

[0019] The product of the cumulative sum of the first average of all adjacent data points of the target data point and the temperature change degree of the target data point is denoted as the fire source feature value of the target data point at the current collection time.

[0020] Further, the specific determination method of the fire source target point and the fire area is:

[0021] The product of the absolute value of the difference between the temperature data of the data points and the Euclidean distance is used as the distance between the data points. All data points in the 3D fire field map at the current collection time are clustered to obtain two data point clusters. All data points in the cluster with the maximum average temperature data are denoted as fire data points. The data point with the maximum fire source feature value in the fire data points is denoted as the fire source target point.

[0022] The region composed of the fire data points is denoted as the fire area.

[0023] Further, the specific method for retaining part of the fire source target point and determining the fire extinguishing quadruped robot for extinguishing the fire source target point according to the number of data points, the temperature change degree and the fire source feature value of the data points in the fire area where the fire source target point is located, and the positional relationship between the fire extinguishing quadruped robot and the fire source target point is:

[0024] When the same fire area contains different fire source target points, the centroid of the fire area is determined, the fire source target point closest to the centroid of the fire area is retained, and other fire source target points in the fire area are deleted.

[0025] The cumulative sum of the product of the temperature change degree and the fire source feature value of all fire data points in the fire area is denoted as the first cumulative sum of the fire area. The product of the first cumulative sum of the fire area and the number of all fire data points in the fire area is denoted as the fire risk degree of the fire source target point in the fire area.

[0026] The Euclidean distance between each fire extinguishing quadruped robot and each fire source target point is calculated, and the fire source target point is extinguished by the fire extinguishing quadruped robot closest to the fire source target point; when the same fire extinguishing quadruped robot extinguishes different fire source target points, the fire extinguishing quadruped robot extinguishes the fire source target points in the order of the fire risk degrees of the fire source target points from large to small.

[0027] Further, the specific acquisition method of the local node path is:

[0028] The initial position of the fire extinguishing quadruped robot in the highway tunnel is set as the root node of the RRT tree, a sampling point is randomly sampled in the 3D fire field map at the current collection moment with a probability of 85% , a random fire source target point is directly taken as the sampling point with a probability of 15% , the node closest to the sampling point in the Euclidean distance is taken as , a step is extended from to to obtain a candidate node , wherein, after a step is extended from to , if the obtained candidate node is outside the line segment determined by and , the candidate node is directly taken as , the path from to is recorded as the local node path from to .

[0029] Further, the determination method of the adjustment step of the fire extinguishing quadruped robot is:

[0030] The Euclidean distance between the fire extinguishing quadruped robot extinguishing the fire source target point and the fire source target point is recorded as the planning distance of the fire extinguishing quadruped robot, the product of the fire risk degree of the fire source target point and the number of obstacles in the local node path is recorded as the second product of the fire source target point, and the normalized value of the product of the second product of the fire source target point and the planning distance of the fire extinguishing quadruped robot extinguishing the fire source target point is recorded as the cost function value of the fire extinguishing quadruped robot;

[0031] When the cost function value of the fire extinguishing quadruped robot is greater than , a new candidate node is randomly searched in the neighborhood radius of , and the value of the cost function of the new candidate node and the local node path from is obtained, until the value of the cost function is less than or equal to , if the value of the cost function cannot meet the condition that the value of the cost function is less than or equal to after 20 times of random search, a sampling point is randomly resampled in the 3D fire field map at the current acquisition moment ; wherein, the third preset threshold is represented by

[0032] When the value of the cost function of the fire extinguishing quadruped robot is less than or equal to , the adjustment step length of the fire extinguishing quadruped robot is determined according to the value of the cost function of the fire extinguishing quadruped robot and the initial step length, and the calculation formula is:

[0033] ;

[0034] wherein, the adjustment step length of the fire extinguishing quadruped robot is represented by the preset initial step length of the RRT algorithm is represented by the value of the cost function of the fire extinguishing quadruped robot is represented by the preset step length adjustment parameter is represented by

[0035] Further, the method for obtaining the fire extinguishing path of the fire extinguishing quadruped robot according to the adjustment step length comprises the following specific method:

[0036] The adjustment step length of the fire extinguishing quadruped robot is taken as the step length value of the fire extinguishing quadruped robot walking towards the fire source target point, a sub-path of the fire extinguishing quadruped robot walking towards the fire source target point is generated, the sub-path is repeatedly determined until the distance between the candidate node and the fire source target point is less than or equal to the preset tolerance length, the generation of the sub-path is stopped, and the path formed by connecting the heads and tails of all the sub-paths is taken as the fire extinguishing path of the fire extinguishing quadruped robot walking towards the fire source target point.

[0037] In the second aspect, the embodiments of the present application also provide a self-operation control system of a highway tunnel fire extinguishing quadruped robot, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the method according to any one of the above embodiments when executing the computer program.

[0038] The present application has the following beneficial effects:

[0039] This application first determines whether a fire has occurred based on a 3D fire scene map of the highway tunnel. Upon detection of a fire, a fire-fighting quadruped robot is activated, and path planning is performed on the robot. Specifically, the severity of temperature changes at target data points is evaluated based on temperature data to assess the likelihood that a target data point is a fire source, i.e., the fire source characteristic value of the target data point at the current acquisition time. Based on the location distribution of all data points and the differences in temperature data on the 3D fire scene map at the current acquisition time, the fire source target point and the fire area are determined. Combining the fire source characteristic value, the fire source target point and the fire-fighting quadruped robot for extinguishing the fire at the target point are determined, where the fire source target point is the center point of the fire. Further, the fire-fighting quadruped robot is positioned according to its location and the location of the fire source target point. The fire extinguishing path is planned. When the probability of collision is high or the possibility of damage due to high temperature is high, the step size of the path planning is reduced. When the fire extinguishing quadruped robot uses local node paths as its actual walking path, when the probability of collision is low and the probability of the fire extinguishing quadruped robot passing through the walking path is high, a larger adjustment step size is selected to further improve the path optimization efficiency and obtain the fire extinguishing path of the fire extinguishing quadruped robot. This solves the problem that the RRT algorithm uses a fixed step size to realize path planning, which cannot meet the needs of different conditions in highway tunnels. It improves the rationality of the path planning of the fire extinguishing quadruped robot in highway tunnels. The fire extinguishing quadruped robot moves according to the fire extinguishing path and reaches the fire extinguishing position. Finally, the fire extinguishing quadruped robot completes the autonomous operation control of fire extinguishing at the fire extinguishing position. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating an autonomous operation control method for a quadruped robot used for firefighting in a highway tunnel, provided in one embodiment of the present invention.

[0042] Figure 2 This is a flowchart illustrating the operation of a fire-extinguishing quadruped robot according to an embodiment of the present invention. Detailed Implementation

[0043] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0044] Please refer to Figure 1 which shows a flow chart of an autonomous operation control method of a highway tunnel fire extinguishing quadruped robot provided by an embodiment of the present application, the method comprising the following steps:

[0045] In step S001, a 3D fire field map and a two-dimensional tunnel map of a highway tunnel at a current collection time and at a previous adjacent collection time are obtained according to a preset sensor in the highway tunnel to be completed with a fire extinguishing task, and a data point in the 3D fire field map corresponds to a spatial position coordinate and temperature data.

[0046] The point cloud data, environmental image data and thermal imaging image data of the highway tunnel are collected using a laser radar, a monitoring camera and a thermal imaging camera arranged in advance in the highway tunnel to be completed with the fire extinguishing task. The point cloud data, environmental image data and thermal imaging image data of the highway tunnel are processed using a SLAM system to obtain the 3D fire field map and the two-dimensional tunnel map of the highway tunnel.

[0047] Among them, the SLAM (Simultaneous Localization and Mapping) system is a simultaneous localization and mapping system; using the SLAM system to obtain the 3D fire field map and the two-dimensional tunnel map is a known technology, and will not be described again. It can be understood that each data point in the 3D fire field map has corresponding spatial position coordinates and temperature data.

[0048] It should be noted that, in order to facilitate calculation, all temperature data involved in the calculation in the present embodiment are subjected to data preprocessing, thereby canceling the dimension effect. In the present embodiment, the Z-Score standard normalization method is used to process the temperature data, and in actual application, other methods such as the existing technology of maximum and minimum value normalization method can be used for de-dimension processing, which is not limited herein.

[0049] Preferably, in an embodiment of the present application, the point cloud data, environmental image data and thermal imaging image data of the highway tunnel are collected uniformly 90 times per minute at the current collection time and before, and the 3D fire field map and the two-dimensional tunnel map of the highway tunnel are obtained synchronously. In actual application, as other implementation manners, the implementer can determine the value of the sampling frequency according to the actual situation, and the present application does not have special limitation.

[0050] At this point, the 3D fire scene map and the two-dimensional tunnel map of the highway tunnel at the current collection time and the adjacent collection time before the current collection time are obtained.

[0051] In step S002, it is determined whether a fire occurs according to the 3D fire scene map of the highway tunnel. If a fire occurs, any one data point in the 3D fire scene map of the highway tunnel is recorded as a target data point. The temperature change degree of the target data point at the current collection time is determined according to the difference between the temperature data of the target data point at the current collection time and the adjacent collection time before the current collection time. The fire source characteristic value of each data point in the 3D fire scene map is determined according to the position of the adjacent data point in the 3D fire scene map at the current collection time, the temperature data of each data point, and the temperature change degree.

[0052] The mean value of the temperature data of all data points in the 3D fire scene map is recorded as a temperature mean value. When there is a data point in the 3D fire scene map whose temperature data is greater than or equal to twice the temperature mean value, it is determined that a fire occurs in the highway tunnel. When the base station arranged inside the tunnel receives the determination instruction that a fire occurs in the highway tunnel, the position of the fire extinguishing quadruped robot is taken as the initial position of the highway tunnel fire extinguishing robot path planning, so as to plan the fire extinguishing path of the highway tunnel fire extinguishing robot.

[0053] When a fire occurs, the temperature at the fire source is the highest, the temperature gradually decreases from the fire source to the surrounding, and the temperature in the area affected by the fire spread is constantly rising, and the temperature change at the fire source is the most violent. Therefore, the temperature change degree of the position of the data point can be evaluated according to the difference between the temperature data of each data point in the 3D fire scene map at the adjacent collection time, and then the target position of the fire extinguishing quadruped robot in the highway tunnel for achieving fire extinguishing is determined.

[0054] Any one data point in the 3D fire scene map of the highway tunnel is recorded as a target data point. The temperature change degree of the target data point at the current collection time is determined according to the difference between the temperature data of the target data point at the current collection time and the adjacent collection time before the current collection time.

[0055] The current collection time and the adjacent collection time before the current collection time are recorded as the target collection time. The absolute value of the difference between the temperature data of the target data point at the target collection time and the adjacent collection time before the target collection time is recorded as the first absolute value of the target data point at the target collection time. The cumulative sum of the first absolute value of all collection times within the current collection time and the adjacent collection time before the current collection time is recorded as the temperature change degree of the target data point at the current collection time.

[0056] It should be noted that, if the current collection time and the adjacent collection time before the current collection time are recorded as the target collection time,​ The data collection time within a minute includes the first data collection time. Since the first data collection time has no preceding adjacent data collection time, the first data collection time is not analyzed when calculating the degree of temperature change. This represents the first preset threshold, which is set to 1 in this embodiment.

[0057] When the target data point is at or before the current acquisition time The greater the difference between temperature data at adjacent acquisition times within a minute, and the greater the temperature data of the target data point at the current acquisition time, the more drastic the temperature change at the target data point location and the higher the temperature, the greater the possibility that the target data point is a source of fire. In this case, the degree of temperature change of the target data point at the current acquisition time is greater.

[0058] Based on the positions of adjacent data points in the 3D fire site map at the current acquisition time, the temperature data of each data point, and the degree of temperature change, the fire source characteristic value of each data point in the 3D fire site map is determined.

[0059] The closest Euclidean distance between the target data point and the 3D fire site map at the current acquisition time. Each data point is designated as an adjacent data point. The direction from the target data point to the adjacent data point is designated as the extension direction of the adjacent data point. For all adjacent data points along the extension direction, the product of the temperature difference between the previous and next adjacent data points and the temperature change of the previous adjacent data point is designated as the first product of the previous adjacent data point. The mean of the first products of all adjacent data points along the extension direction is designated as the first mean of the adjacent data points. The sum of the first means of all adjacent data points of the target data point and the product of the temperature change of the target data point are designated as the fire source characteristic value of the target data point at the current acquisition time.

[0060] It is important to note that, in the extension direction of adjacent data points, if there are no other adjacent data points after the last adjacent data point in the extension direction, this data point is not analyzed when calculating the first product; This represents the second preset threshold, which is set to 20 in this embodiment.

[0061] The greater the first product of all adjacent data points in the extension direction of the adjacent data points of the target data point, the more obvious the feature of the target data point position gradually decreasing from the fire source to the surrounding along the extension direction of the adjacent data points, and the greater the temperature change degree of the target data point, the more intense the temperature data change of the target data point position and the higher the temperature, the greater the possibility of the target data point being the fire source, and the greater the fire source feature value of the target data point.

[0062] At this point, the fire source feature value of the target data point at the current acquisition moment is obtained.

[0063] Step S003, according to the difference between the position distribution and the temperature data of all data points in the 3D fire field map at the current acquisition moment, the fire source target point and the fire area are determined, according to the number of data points, the temperature change degree and the fire source feature value of the data points in the fire area where the fire source target point is located, and the positional relationship between the fire extinguishing quadruped robot and the fire source target point, part of the fire source target points are reserved, and the fire extinguishing quadruped robot for extinguishing the fire source target point is determined.

[0064] The data points in the 3D fire field map at the current acquisition moment are clustered, the distance between the data points is the product of the absolute value of the difference between the temperature data of the data points and the Euclidean distance, two data point clusters are obtained, all data points in the cluster with the maximum mean value of temperature data are recorded as fire data points, and the data point with the maximum fire source feature value in the fire data points is recorded as the fire source target point. The fire data points are analyzed by connected domain analysis to obtain the fire area. That is, the region composed of the fire data points is recorded as the fire area.

[0065] The fire source target point is the fire source point determined before the fire extinguishing quadruped robot starts path planning. Since the quadruped robot can only process one fire source point at a time when extinguishing the fire, there may be multiple fire source points in the highway tunnel, so when the number of fire source target points is greater than 1, the order of extinguishing all fire source target points needs to be sorted.

[0066] When the same fire area contains different fire source target points, the centroid of the fire area is determined, only the fire source target point closest to the centroid of the fire area in the Euclidean distance is retained, and other fire source target points in the fire area are deleted.

[0067] At this point, different fire source target points are in different fire areas.

[0068] For different fire source target points in different fire areas, the fire risk degree of the fire source target point in the fire area is determined according to the number of fire data points, the temperature change degree and the fire source feature value in the fire area.

[0069] The accumulated sum of the product of the temperature variation degree of all fire data points in the fire area and the fire source characteristic value is denoted as a first accumulated sum of the fire area, and the product of the first accumulated sum of the fire area and the number of all fire data points in the fire area is denoted as a fire risk degree of the fire source target point in the fire area.

[0070] The greater the fire risk degree of the fire source target point is, the more the fire source target point should be prioritized for fire extinguishing.

[0071] The Euclidean distance between each fire extinguishing quadruped robot and each fire source target point is calculated, the fire extinguishing is implemented on the fire source target point by the fire extinguishing quadruped robot closest to the fire source target point, and when the same fire extinguishing quadruped robot extinguishes different fire source target points, the fire extinguishing quadruped robot extinguishes each fire source target point in order from large to small according to the fire risk degree of the fire source target point.

[0072] Thus, the fire source target point and the fire extinguishing quadruped robot for extinguishing the fire source target point are determined.

[0073] In step S004, the initial value of the parameter of the RRT algorithm is set according to the position of the fire extinguishing quadruped robot, the local node path is obtained, the adjustment step of the fire extinguishing quadruped robot is determined in combination with the positional relationship between the fire extinguishing quadruped robot and the fire source target point, the fire extinguishing path of the fire extinguishing quadruped robot is obtained according to the adjustment step, and the fire extinguishing quadruped robot travels according to the fire extinguishing path to reach the fire extinguishing position.

[0074] The fire extinguishing quadruped robot for extinguishing each fire source target point is path planned. Specifically, as an embodiment of the present application, the initial step of the RRT algorithm is set to one fourth of the minimum value of the width in the highway tunnel, the maximum iteration number is set to 10,000, the neighborhood radius is set to 4 times the step, the initial position of the fire extinguishing quadruped robot in the highway tunnel is set to the root node of the RRT tree, a sampling point is randomly sampled in the 3D fire field map at the current collection moment with a probability of 85% a random fire source target point is directly taken as the sampling point with a probability of 15% the node closest to the sampling point in the Euclidean distance is taken as a step is extended from to to obtain a candidate node , wherein, after a step is extended from to , if the obtained candidate node is outside the line segment determined by and , then is directly taken as the candidate node , and is connected to The path, denoted as arrive The local node path.

[0075] The Euclidean distance between the fire-fighting quadruped robot and the fire source target point is denoted as the planned distance of the fire-fighting quadruped robot. The product of the fire risk level of the fire source target point and the number of obstacles in the local node path is denoted as the second product of the fire source target point. The normalized value of the product of the second product of the fire source target point and the planned distance of the fire-fighting quadruped robot is denoted as the cost function value of the fire-fighting quadruped robot.

[0076] The number of obstacles in the local node path can be determined based on the 3D fire map of the highway tunnel.

[0077] When the cost function value of the fire-fighting quadruped robot is greater than At that time, the cost of the firefighting quadruped robot using local node paths for advancement was too high. Randomly search for new candidate nodes within the neighborhood radius, and obtain the new candidate nodes and... The value of the cost function for the local node path is determined until the value of the cost function is less than or equal to the value of the path. If after 20 random searches, the cost function value still cannot be less than or equal to... Then, random sampling points will be resampled on the 3D fire site map at the current sampling time. .

[0078] in, This represents the third preset threshold, which is set to 0.8 in this embodiment.

[0079] When the cost function value of the fire-fighting quadruped robot is less than or equal to At that time, based on the cost function value and initial step size of the fire-fighting quadruped robot, the adjustment step size of the fire-fighting quadruped robot is determined:

[0080] ;

[0081] in, This indicates the adjustment step size of the firefighting quadruped robot; This represents the initial step size of the RRT algorithm. In this embodiment, the initial step size is set to one-quarter of the minimum width inside the highway tunnel. This represents the cost function value of the firefighting quadruped robot; This represents the preset step size adjustment parameter. In this embodiment, the step size adjustment parameter is set to 0.1. The function of the step size adjustment parameter is to adjust the specific value of the step size of the fire-fighting quadruped robot.

[0082] When the cost function value of the fire extinguishing quadruped robot is larger, the possibility of collision and damage caused by high temperature is greater when the fire extinguishing quadruped robot walks along the local node path as the actual walking path, and at this time, a smaller adjustment step value should be selected; when the cost function value of the fire extinguishing quadruped robot is smaller, the possibility of collision is smaller and the possibility of the fire extinguishing quadruped robot passing through the walking path is greater when the fire extinguishing quadruped robot walks along the local node path as the actual walking path, and at this time, a larger adjustment step value should be selected to further improve the path optimization efficiency.

[0083] The adjustment step of the fire extinguishing quadruped robot is taken as the step value of the fire extinguishing quadruped robot walking to the fire source target point, and a sub-path of the fire extinguishing quadruped robot walking to the fire source target point is generated, and the sub-path is repeatedly determined until the distance between the candidate node The fire extinguishing quadruped robot according to the fire extinguishing path advances to the fire source target point, and advances to a position where the distance between the position and the fire source target point is less than or equal to the tolerance length, and the position of the fire extinguishing quadruped robot is recorded as the fire extinguishing position.

[0084] The tolerance length is a preset length, and the value of the tolerance length in the embodiment is 2 meters.

[0085] At this point, the fire extinguishing quadruped robot advances according to the fire extinguishing path, and reaches a position where the distance between the position and the fire source target point is less than or equal to the tolerance length, that is, reaches the fire extinguishing position.

[0086] Step S005, the fire extinguishing quadruped robot completes the autonomous operation control of fire extinguishing at the fire extinguishing position.

[0087] The fire extinguishing quadruped robot performs 360-degree surround scanning on the fire source of the fire source target point through the built-in high-definition thermal imaging and visible light camera at the fire extinguishing position, obtains the fire field surround image of the fire extinguishing quadruped robot at the fire extinguishing position, determines the accurate position of the fire hydrant and the flame, the spreading direction, and realizes the fire type identification, at the same time, identifies the explosion hazard sources such as oil tank, pressure container and combustible material, and evaluates the development stage of the fire and the fire source.

[0088] Wherein, the development stage of the fire source includes three stages of initial, development and fierce; the fire type is divided into A class fire and B class fire, wherein, the interior materials such as seats, carpets, plastic parts of trucks, buses and other vehicles are on fire, the ordinary goods such as wood, paper, textiles, plastic products transported by vehicles are on fire, and the combustible solids such as cable insulation layer, maintenance materials in the tunnel are on fire, which all belong to A class fire, and can be directly connected and utilized by using the fire fighting system to extinguish the fire; B class fire includes the combustible liquid such as gasoline, diesel, oil, lubricating oil leaked from oil tank truck, fuel truck, car fuel tank and other vehicles, the combustible liquid or meltable solid such as asphalt, certain chemical products leaked from the vehicles, the initial fire of the live equipment such as distribution box, lighting facilities, vehicle electrical system in the tunnel, and the fire of the water reaction materials such as metallic sodium, potassium and certain chemicals, and the water extinguishing of B class fire will cause explosion or produce toxic gas.

[0089] According to the surrounding image of the fire scene, the accurate position of the fire hydrant and the flame, the spreading direction, the fire type recognition, the identification of the explosion hazard sources and the burning materials such as oil tank and pressure container, and the assessment of the fire and the development stage of the fire source are all known technologies, and will not be described in detail.

[0090] The fire extinguishing quadruped robot generates a fire source target point fire extinguishing scheme according to the preset rules of the fire extinguishing decision, and transmits the fire extinguishing scheme back to the tunnel central monitoring system for confirmation by the operator.

[0091] For A class fire, the fire extinguishing quadruped robot determines the best fire extinguishing point according to the scheme confirmed by the operator, reaches the best fire extinguishing point, realizes the full-automatic and rapid docking of the fire fighting water supply hose and the fire hydrant, completes the start of the fire extinguishing device of the fire extinguishing quadruped robot, and automatically adjusts the pitch angle and azimuth angle of the fire extinguishing device of the fire extinguishing quadruped robot according to the guidance of the thermal imaging camera, accurately aims at the flame root or the area to be covered, until the temperature of the fire source target point is less than the safe temperature and there is no visible flame, and the fire control is realized. The fire extinguishing quadruped robot transmits the state of "realizing the fire control" to the tunnel central monitoring system.

[0092] Wherein, the preset rules of the fire extinguishing decision give priority to cooling the oil tank and avoiding the water flow from hitting the electrical equipment, which are preset by the person skilled in the art; the fire extinguishing scheme includes covering the flame root with dry powder extinguishing agent and paying attention to cooling the oil tank, which are preset by the person skilled in the art; the fire extinguishing of the fire extinguishing quadruped robot at the best fire extinguishing point can maintain a stable standing posture and effectively cover the fire source, while avoiding the main heat radiation and potential explosion risk; the safe temperature is a parameter preset by the person skilled in the art, and in this embodiment, the safe temperature is set to 60℃.

[0093] For Class B fires, the fire-fighting quadruped robot determines the optimal extinguishing point based on the operator-confirmed plan. The valve of the dry powder extinguishing agent storage tank on the robot's back or midsection opens, and the high-precision pan-tilt fire cannon integrated into the robot's torso automatically adjusts its elevation and azimuth angles, guided by a thermal imager, precisely targeting the base of the flames or areas requiring coverage. High-pressure nitrogen drives the dry powder extinguishing agent, which is then sprayed at high speed through the nozzle, forming a dense powder mist that covers the flames. The quadruped robot can make short, small movements or adjust the spray angle to ensure effective coverage of the entire fire source area, especially hard-to-reach corners. During spraying, the thermal imager continuously monitors the flame temperature and range changes until the temperature at the target fire point is below a safe level and there are no visible flames, achieving fire control. The quadruped robot transmits the "fire control achieved" status to the tunnel's central monitoring system.

[0094] The workflow diagram of the fire-fighting quadruped robot is as follows: Figure 2 As shown. In Figure 2 In this process, the environmental perception of the highway tunnel corresponds to the determination of whether a fire has occurred in steps S001 and S002. When the fire-fighting quadruped robot receives the determination command that "a fire has occurred in the highway tunnel", it starts and plans the fire-fighting path. In the process of planning the fire-fighting path, it is necessary to locate the fire source, which means determining the target point of the fire source. Acquiring the temperature level data model is the process of determining the fire risk level. Constructing a priority queue is the process of determining the fire-fighting quadruped robot to extinguish the fire source target point based on the fire risk level. RTT algorithm optimization is the process of determining the fire-fighting path. Feedback on the fire situation at the tunnel site is the process of the fire-fighting quadruped robot completing the autonomous operation control of the fire-fighting at the fire-fighting position. When the fire-fighting quadruped robot transmits the status of "achieving fire control" to the central monitoring system of the tunnel, the fire-fighting quadruped robot issues a retreat command and returns to the initial position.

[0095] This completes the autonomous operation control of the quadruped robot for firefighting in highway tunnels.

[0096] Based on the same inventive concept as the above method, this embodiment of the invention also provides an autonomous operation control system for a quadruped robot for firefighting in highway tunnels, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described autonomous operation control methods for a quadruped robot for firefighting in highway tunnels.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An autonomous operation control method for a quadruped robot used for firefighting in highway tunnels, characterized in that, The method includes the following steps: Based on the sensors pre-set in the highway tunnel where the fire extinguishing task is to be completed, a 3D fire scene map and a 2D tunnel map of the highway tunnel at the current collection time and at different adjacent collection times are obtained. The data points in the 3D fire scene map correspond to the spatial location coordinates and temperature data. The 3D fire map of the highway tunnel is used to determine whether a fire has occurred. If a fire has occurred, any data point in the 3D fire map of the highway tunnel is recorded as a target data point. The temperature change of the target data point at the current acquisition time is determined based on the difference between the temperature data of the target data point at the current acquisition time and the temperature change at the previous adjacent acquisition time. Based on the position of the adjacent data points in the 3D fire map at the current acquisition time, the temperature data of each data point and the temperature change, the fire source characteristic value of each data point in the 3D fire map is determined. Based on the location distribution of all data points and the differences between temperature data in the 3D fire site map at the current acquisition time, the fire source target point and the fire area are determined. Based on the number of data points, the degree of temperature change and the fire source characteristic value in the fire area where the fire source target point is located, as well as the positional relationship between the fire-fighting quadruped robot and the fire source target point, some fire source target points are retained, and the fire-fighting quadruped robot to extinguish the fire source target point is determined. The initial values ​​of the RRT algorithm parameters are set according to the position of the fire-fighting quadruped robot. The local node path is obtained. The adjustment step size of the fire-fighting quadruped robot is determined by combining the positional relationship between the fire-fighting quadruped robot and the target point of the fire source. The fire-fighting path of the fire-fighting quadruped robot is obtained according to the adjustment step size. The fire-fighting quadruped robot moves according to the fire-fighting path and reaches the fire-fighting position. The firefighting quadruped robot performs autonomous firefighting operations at the fire location; The method for determining the step length of the fire-fighting quadruped robot is as follows: The initial position of the fire-fighting quadruped robot inside the highway tunnel is set as the root node of the RRT tree, and a sampling point is randomly sampled from the 3D fire site map at the current acquisition time with an 85% probability. With a 15% probability, a random fire source target point is directly used as the sampling point. , will be with sampling points The node with the closest Euclidean distance is used as ,from Towards Extend the direction by one step to obtain candidate nodes. ; The Euclidean distance between the fire-fighting quadruped robot and the fire source target point is denoted as the planned distance of the fire-fighting quadruped robot. The product of the fire risk level of the fire source target point and the number of obstacles in the local node path is denoted as the second product of the fire source target point. The normalized value of the product of the second product of the fire source target point and the planned distance of the fire-fighting quadruped robot is denoted as the cost function value of the fire-fighting quadruped robot. When the cost function value of the fire-fighting quadruped robot is greater than At that time, Randomly search for new candidate nodes within the neighborhood radius, and obtain the new candidate nodes and... The value of the cost function for the local node path is determined until the value of the cost function is less than or equal to the value of the path. If after 20 random searches, the cost function value still cannot be less than or equal to... Then, random sampling points will be resampled on the 3D fire site map at the current sampling time. ;in, This indicates the third preset threshold. When the cost function value of the fire-fighting quadruped robot is less than or equal to At that time, based on the cost function value and initial step size of the fire-fighting quadruped robot, the adjustment step size of the fire-fighting quadruped robot is determined, and the calculation formula is as follows: ； in, This indicates the adjustment step size of the firefighting quadruped robot; This indicates the preset initial step size of the RRT algorithm; This represents the cost function value of the firefighting quadruped robot; This indicates the preset step size adjustment parameter; The specific method for obtaining the fire extinguishing path of the fire-extinguishing quadruped robot based on the adjusted step size is as follows: The adjusted step size of the fire-fighting quadruped robot is used as the step size value for the robot to walk towards the fire source target point. A sub-path for the fire-fighting quadruped robot to walk towards the fire source target point is generated. The sub-path is repeatedly determined until a candidate node is reached. If the distance between the target point and the fire source is less than or equal to the preset tolerance length, stop generating sub-paths, and connect all the sub-paths end to end to form the path for the fire-extinguishing quadruped robot to walk towards the target point of the fire source.

2. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The specific methods for determining whether a fire has occurred based on a 3D fire scene map of a highway tunnel include: The average temperature of all data points in the 3D fire map is recorded as the average temperature. When the temperature of a data point in the 3D fire map is greater than or equal to twice the average temperature, it is determined that a fire has occurred in the highway tunnel.

3. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The degree of temperature change of the target data point at the current acquisition time is determined as follows: The current collection time and previous times Any acquisition time within a minute is recorded as the target acquisition time. The absolute value of the difference between the temperature data of the target data point at the target acquisition time and the temperature data of the previous adjacent acquisition time is recorded as the first absolute value of the target data point at the target acquisition time. The target data points are at or before the current acquisition time. The sum of the first absolute values ​​of all data collected within a minute, multiplied by the temperature data of the target data point at the current data collection time, is denoted as the degree of temperature change of the target data point at the current data collection time; where... This indicates the first preset threshold.

4. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The specific method for obtaining the fire source feature value of each data point in the 3D fire site map is as follows: The closest Euclidean distance between the target data point and the 3D fire site map at the current acquisition time. A number of data points are designated as adjacent data points. The direction from the target data point to the adjacent data points is designated as the extension direction of the adjacent data points. For all adjacent data points along the extension direction, the product of the difference in temperature data between the preceding and following adjacent data points and the degree of temperature change of the preceding adjacent data point is designated as the first product of the preceding adjacent data points. This indicates the second preset threshold. The mean of the first product of all adjacent data points along the extension direction of adjacent data points is denoted as the first mean of adjacent data points. The sum of the first means of all adjacent data points of the target data point and the product of the temperature change of the target data point are recorded as the fire source characteristic value of the target data point at the current acquisition time.

5. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The specific method for determining the target point of the fire source and the fire area is as follows: The product of the absolute value of the temperature data difference between data points and the Euclidean distance is used as the distance between data points. All data points in the 3D fire site map at the current acquisition time are clustered to obtain two data point clusters. All data points in the cluster with the largest mean temperature data are recorded as fire data points, and the data point with the largest fire source feature value among the fire data points is recorded as the fire source target point. The area comprised of fire data points is designated as the fire zone.

6. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The specific method for determining which fire-fighting quadruped robot to extinguish fires at target fire points, based on the number of data points within the fire area where the target fire point is located, the degree of temperature change, and the characteristic values ​​of the fire source, as well as the positional relationship between the fire-fighting quadruped robot and the target fire point, includes retaining a portion of the target fire points and determining the fire-fighting quadruped robot to extinguish the fires at the target fire points. When the same fire area contains different fire source target points, determine the centroid of the fire area, retain the fire source target point with the closest Euclidean distance to the centroid of the fire area, and delete other fire source target points in the fire area. The sum of the products of the temperature change degree and the fire source characteristic value of all fire data points in the fire area is recorded as the first sum of the fire area. The product of the first sum of the fire area and the number of all fire data points in the fire area is recorded as the fire risk degree of the fire source target point in the fire area. Calculate the Euclidean distance between each fire-fighting quadruped robot and each fire source target point. The fire-fighting quadruped robot closest to the fire source target point will extinguish the fire at the fire source target point. When the same fire-fighting quadruped robot extinguishes fires at different fire source target points, the fire-fighting quadruped robot will extinguish the fires at the fire source target points in descending order of the fire risk level corresponding to the fire source target points.

7. The autonomous operation control method for a quadruped robot for fire extinguishing in highway tunnels according to claim 1, characterized in that, The specific method for obtaining the local node path is as follows: from Towards If the candidate nodes are obtained after extending the direction by one step... exist and Outside of the defined line segments, directly... as a candidate node ,Will arrive The path, denoted as arrive The local node path.

8. An autonomous operation control system for a quadruped robot used for firefighting in highway tunnels, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-7.

Images