Roadway gas environment inspection method based on coal mine inspection robot and related equipment

Abstract

The invention relates to the field of coal mine roadway detection, and discloses a roadway gas environment inspection method and device based on a coal mine inspection robot and related equipment. And setting a detection step length, and determining an observation point and an observation area surface according to the roadway length and the detection step length. And controlling the inspection robot to move along the mine roadway, and collecting gas environment data. And taking data acquired in the process that the inspection robot moves from the (i-1) th observation area surface to the ith observation area surface as a target gas environment data set. And calculating the target gas environment data set by adopting a DPC clustering algorithm, and outputting an (i-1) th clustering center. And constructing a current space initial gas diffusion model according to the gas environment data corresponding to the (i-1) th clustering center, and calculating a corresponding gas concentration calculation value. And if the difference value between the gas concentration calculation value and the gas concentration sampling value meets a preset condition, generating a gas distribution condition of the (i-1) th sampling space, and further generating a coal mine tunnel gas distribution diagram according to the gas distribution condition of each sampling space.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine tunnel detection, and in particular to a tunnel gas environment inspection method based on a coal mine inspection robot and related equipment. Background Art

[0002] Coal mine tunnels are high-risk working environments, and gas safety monitoring is directly related to the lives and production safety of miners. The dynamic diffusion of harmful gases such as gas and carbon monoxide in the tunnels can easily lead to localized accumulation. Traditional fixed sensors have insufficient spatial coverage, making it difficult to capture changes in gas concentration gradients in real time, especially under complex ventilation conditions, which can easily lead to monitoring blind spots. Although existing mobile inspection technologies use robots equipped with multi-parameter sensors to increase the detection range, the dimensional differences and noise interference of multi-source heterogeneous data (such as gas concentration, temperature and humidity, wind speed, and tunnel spatial parameters) make data fusion difficult, affecting the iterative accuracy and timeliness of gas diffusion models.

[0003] Current data processing methods often rely on fixed thresholds or static clustering algorithms, which are difficult to adapt to the dynamic characteristics of roadway gas distribution. Traditional algorithms lack compatibility with multi-dimensional parameters, and manual intervention easily introduces subjective bias, resulting in delayed model optimization. Gas diffusion models lack a good match with the real environment, limiting the accuracy of hazardous area identification and the effectiveness of early warnings. An adaptive data fusion and dynamic modeling approach is urgently needed for precise monitoring. Summary of the Invention

[0004] The invention discloses a mine inspection robot-based method, device and related equipment for inspecting the gas environment in a tunnel.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a method for inspecting the gas environment in a coal mine tunnel based on a coal mine inspection robot, which is applied to an inspection robot performing inspections in a coal mine tunnel. The method includes:

[0007] The inspection step length of the inspection robot is set, and i observation points and i observation area surfaces are determined according to the lane length and the inspection step length; wherein the observation points and the observation area surfaces correspond one to one, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; and the distance between the i-1-th observation area surface and the i-th observation area surface is the inspection step length;

[0008] Controlling the inspection robot to move along the lane, and collecting gas environment data during the movement of the inspection robot;

[0009] The target gas environment data collected during the inspection robot's movement from the (i-1)th observation area to the (i)th observation area is used as a target gas environment data set;

[0010] The DPC clustering algorithm is used to calculate the target gas environment data set and output the i-1th cluster center;

[0011] The target gas environment data corresponding to the i-1th cluster center is used as the input parameter for optimizing the parameters of the coal mine tunnel gas diffusion model, and the current space initial gas diffusion model is constructed based on the particle swarm optimization algorithm;

[0012] Using the current spatial initial gas diffusion model, a calculated gas concentration value corresponding to the i-1th cluster center is obtained;

[0013] If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center meets the preset conditions, the gas distribution of the i-1th sampling space is generated using the current spatial initial gas diffusion model; the i-1th sampling space is the sampling space between the i-1th observation area surface and the i-th observation area surface;

[0014] If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center does not meet the preset condition, the current spatial initial gas diffusion model is rebuilt until the difference between the calculated gas concentration value output by the current spatial initial gas diffusion model and the corresponding sampled gas concentration value meets the preset condition;

[0015] After obtaining the gas distribution of each sampling space, a coal mine roadway gas distribution map is generated according to the gas distribution of each sampling space.

[0016] In one embodiment, the DPC clustering algorithm relative distance calculation method adopts the Mahalanobis distance calculation method.

[0017] In one embodiment, the Mahalanobis distance calculation method is:

[0018] Calculate the Mahalanobis distance between each target gas environment data in the target gas environment data set and the target gas environment data set:

[0019]

[0020] Wherein, μ is the overall mean of the target gas environment data in the target gas environment data set; V is the covariance matrix of the target gas environment data set, and x is the target gas environment data;

[0021] The covariance matrix is:

[0022]

[0023] Wherein, n is the total number of target gas environment data in the target gas environment data set, m means that each target gas environment data is a row vector containing m variables; v ij Represents the covariance of any two samples Xi and Xj in the same data set;

[0024] v ij =Cov(X i ,X j ),i,j=1,2,...n

[0025] In the target gas environment data set, the Mahalanobis distance between the xth target gas environment data and the yth target gas environment data is:

[0026]

[0027] Wherein, x and y are two different target gas environment data in the target gas environment data set.

[0028] In one embodiment, the DPC clustering algorithm is used to calculate the target gas environment data set and output the i-1th cluster center, including:

[0029] Calculate the cluster center determination parameter of each target gas environment data in the target gas environment data set:

[0030] γ i =P i ·Δ i ;

[0031] Among them, γ i is the cluster center determination parameter of the i-th target gas environment data, P i is the local density ρ of the i-th target gas environment data i The normalized value, Δ i is the relative high density point distance δ to the i-th target gas environment data i Normalized value;

[0032]

[0033] Among them, ρ min is ρ i The minimum value in ρ max is ρ i The maximum value in ;

[0034]

[0035] Among them, δ max is δ i The minimum value in δ min is δ i The maximum value in ;

[0036] The target gas environment data with the largest cluster center determination parameter in the target gas environment data set is used as the target gas environment data corresponding to the i-1th cluster center.

[0037] In one embodiment, before setting the detection step length of the inspection robot, the method further includes:

[0038] Controlling the inspection robot to move in the coal mine tunnel to collect gas concentration and main airflow direction;

[0039] Keeping other variables constant, the partial derivative of the main wind flow direction is calculated to obtain the gas concentration change rate; the gas concentration change rate is used to represent the degree of change in gas concentration when moving along the main wind flow direction.

[0040] In one embodiment, setting the detection step of the inspection robot includes:

[0041] The detection step length of the inspection robot is determined according to the gas concentration change rate; within the detection step length, the fluctuation of the gas concentration change rate is less than a threshold value.

[0042] In one embodiment, the target gas environment data includes the tunnel size, wind speed, and concentrations of multiple gases collected by the inspection robot at the current detection point.

[0043] In a second aspect, an embodiment of the present application provides a tunnel gas environment inspection device based on a coal mine inspection robot, the device comprising:

[0044] A setting module is used to set the detection step length of the inspection robot and determine i observation points and i observation area surfaces based on the lane length and the detection step length; wherein the observation points and observation area surfaces have a one-to-one correspondence, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; and the distance between the i-1-th observation area surface and the i-th observation area surface is the detection step length;

[0045] an acquisition module, configured to control the inspection robot to move along the lane, and collect gas environment data during the movement of the inspection robot; and to use the target gas environment data collected during the movement of the inspection robot from the i-1th observation area surface to the i-th observation area surface as the target gas environment data dataset;

[0046] A calculation module, configured to calculate the target gas environment data set using a DPC clustering algorithm and output an i-1th cluster center;

[0047] A construction module is used to use the target gas environment data corresponding to the i-1th cluster center as input parameters for optimizing the parameters of the coal mine tunnel gas diffusion model, and to construct an initial gas diffusion model of the current space based on a particle swarm optimization algorithm;

[0048] An analysis module, configured to obtain a calculated gas concentration value corresponding to the i-1th cluster center using a current spatial initial gas diffusion model;

[0049] An output module, configured to generate the gas distribution of the i-1 sampling space using the initial gas diffusion model of the current space if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center meets a preset condition; the i-1 sampling space is the sampling space between the i-1 observation area surface and the i-th observation area surface; if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center does not meet the preset condition, reconstruct the initial gas diffusion model of the current space until the difference between the calculated gas concentration value output by the initial gas diffusion model of the current space and the corresponding sampled gas concentration value meets the preset condition;

[0050] The generation module is used to generate a coal mine roadway gas distribution map according to the gas distribution of each sampling space after obtaining the gas distribution of each sampling space.

[0051] In one embodiment, the DPC clustering algorithm relative distance calculation method adopts the Mahalanobis distance calculation method.

[0052] In one embodiment, the Mahalanobis distance calculation method is:

[0053] Calculate the Mahalanobis distance between each target gas environment data in the target gas environment data set and the target gas environment data set:

[0054]

[0055] Wherein, μ is the overall mean of the target gas environment data in the target gas environment data set; V is the covariance matrix of the target gas environment data set, and x is the target gas environment data;

[0056] The covariance matrix is:

[0057]

[0058] Wherein, n is the total number of target gas environment data in the target gas environment data set, m means that each target gas environment data is a row vector containing m variables, v ij Represents the covariance of any two samples Xi and Xj in the same data set;

[0059] v ij =Cov(X i ,X j ),i,j=1,2,...n;

[0060] In the target gas environment data set, the Mahalanobis distance between the xth target gas environment data and the yth target gas environment data is:

[0061]

[0062] Wherein, x and y are two different target gas environment data in the target gas environment data set.

[0063] In one embodiment, the building block is specifically configured to:

[0064] Calculate the cluster center determination parameter of each target gas environment data in the target gas environment data set:

[0065] γ i =P i ·Δ i ;

[0066] Among them, γ i is the cluster center determination parameter of the i-th target gas environment data, P i is the local density ρ of the i-th target gas environment data i The normalized value, Δ i is the relative high density point distance δ to the i-th target gas environment data i Normalized value;

[0067]

[0068] Among them, ρ min is ρ i The minimum value in ρ max is ρ i The maximum value in ;

[0069]

[0070] Among them, δ max is δ i The minimum value in δ min is δ i The maximum value in ;

[0071] The target gas environment data with the largest cluster center determination parameter in the target gas environment data set is used as the target gas environment data corresponding to the i-1th cluster center.

[0072] In one embodiment, the acquisition module is further configured to:

[0073] Controlling the inspection robot to move in the coal mine tunnel to collect gas concentration and main airflow direction;

[0074] Keeping other variables constant, the partial derivative of the main wind flow direction is calculated to obtain the gas concentration change rate; the gas concentration change rate is used to represent the degree of change in gas concentration when moving along the main wind flow direction.

[0075] In one embodiment, the setting module is specifically configured to:

[0076] The detection step length of the inspection robot is determined according to the gas concentration change rate; within the detection step length, the fluctuation of the gas concentration change rate is less than a threshold value.

[0077] In one embodiment, the target gas environment data includes the tunnel size, wind speed, and concentrations of multiple gases collected by the inspection robot at the current detection point.

[0078] In a third aspect, an embodiment of the present application provides a control device comprising a processor and a memory, wherein the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to complete the tunnel gas environment inspection method based on a coal mine inspection robot as described in any one of the first aspects.

[0079] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing a computer program, which is loaded by a processor to execute the tunnel gas environment inspection method based on a coal mine inspection robot as described in any one of the first aspects.

[0080] Compared with the existing technology, the method, device and related equipment for inspecting the gas environment in a roadway based on a coal mine inspection robot provided by the present invention have the following technical effects:

[0081] By constructing a "concentration gradient-detection step" dynamic response mechanism, an intelligent upgrade of the coal mine tunnel gas detection strategy has been achieved. Its core effect is to convert the physical laws of gas diffusion (such as the concentration gradient change rate) into quantifiable detection control parameters, so that the spatial sampling density of the inspection robot can autonomously adapt to the risk distribution characteristics of the tunnel environment. The dynamic detection data is deeply coupled with the gas diffusion model, and the multi-dimensional data features are extracted through the improved DPC clustering algorithm to drive the iterative optimization of the model parameters, and finally a gas distribution calculation system with spatiotemporal adaptability is formed. It solves the problem of model distortion caused by environmental response hysteresis and provides a highly reliable dynamic modeling foundation for coal mine tunnel hazardous gas early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1A schematic diagram of a method for inspecting the gas environment in a mine tunnel using a coal mine inspection robot according to an embodiment of the present invention;

[0083] Figure 2a A schematic diagram of a first gas concentration change rate provided by an embodiment of the present invention;

[0084] Figure 2b A schematic diagram of a second gas concentration change rate provided by an embodiment of the present invention;

[0085] Figure 2c A schematic diagram of a third gas concentration change rate provided by an embodiment of the present invention;

[0086] Figure 2d A schematic diagram of a fourth gas concentration change rate provided by an embodiment of the present invention;

[0087] Figure 2e A schematic diagram of a fifth gas concentration change rate provided by an embodiment of the present invention;

[0088] Figure 2f A schematic diagram of a sixth gas concentration change rate provided by an embodiment of the present invention;

[0089] Figure 2g A schematic diagram of a seventh gas concentration change rate provided by an embodiment of the present invention;

[0090] Figure 2h A schematic diagram of the eighth gas concentration change rate provided by an embodiment of the present invention;

[0091] Figure 2i A schematic diagram of a ninth gas concentration change rate provided in an embodiment of the present invention;

[0092] Figure 2j A schematic diagram of a tenth gas concentration change rate provided by an embodiment of the present invention;

[0093] Figure 2k A schematic diagram of an eleventh gas concentration change rate provided by an embodiment of the present invention;

[0094] Figure 2l A schematic diagram of a twelfth gas concentration change rate provided in an embodiment of the present invention;

[0095] Figure 3 A schematic diagram of a gas monitoring area provided by an embodiment of the present invention;

[0096] Figure 4 A schematic diagram of an observation point and observation area provided by an embodiment of the present invention;

[0097] Figure 5 A schematic structural diagram of an inspection robot provided by an embodiment of the present invention;

[0098] Figure 6 A schematic diagram of a movable bracket for installing an experimental measuring point in a simulated tunnel provided by an embodiment of the present invention;

[0099] Figure 7 A graph showing raw gas detection data from an inspection robot according to an embodiment of the present invention;

[0100] Figure 8 A data curve graph after processing by an improved DPC clustering algorithm provided by an embodiment of the present invention;

[0101] Figure 9 A PSO fitness function diagram provided by an embodiment of the present invention;

[0102] Figure 10 A graph showing the calculated and experimental gas concentration values provided by an embodiment of the present invention;

[0103] Figure 11 A comparison chart of calculated and measured values provided by an embodiment of the present invention;

[0104] Figure 12 A gas concentration distribution diagram of a target observation point M provided in an embodiment of the present invention.

[0105] Reference numerals:

[0106] In the figure, bracket one 1, bracket two 2, bracket three 3, and sensor 4. DETAILED DESCRIPTION

[0107] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0108] The present invention discloses a method for inspecting the gas environment in a coal mine tunnel based on an inspection robot, which is applied to an inspection robot that inspects in a coal mine tunnel. Figure 1 As shown, the method includes:

[0109] S101: Setting the detection step length of the inspection robot, and determining i observation points and i observation area surfaces according to the tunnel length and the detection step length.

[0110] Among them, the observation points and observation area surfaces correspond one to one, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; the distance between the i-1-th observation area surface and the i-th observation area surface is the detection step length.

[0111] The detection step size is the preset spacing between adjacent observation areas as the inspection robot moves within the tunnel. Its physical meaning is the observable spatial scale of gas diffusion characteristics along the tunnel axis. This parameter must be dynamically adjustable, meaning it must be adjusted in real time based on tunnel ventilation parameters (wind speed, gas outflow rate) to ensure spatial continuity between adjacent observation areas.

[0112] The tunnel length refers to the total extension length of the tunnel to be inspected, which is used to determine the number i of observation points and observation area surfaces in combination with the detection step length.

[0113] The observation point is any spatial coordinate point located within the i-th observation area (usually the geometric center point of the area is selected) and serves as the logical identification point of the area.

[0114] The observation area is a continuous spatial slice extending along the axial direction of the roadway, centered at the i-th observation point. The axial length of the observation area is equal to the detection step size, and the radial range of the observation area covers the entire cross-section of the roadway (the width of the observation area matches the roadway width).

[0115] Each observation area corresponds to a unique observation point, and the two form a "point-surface" mapping relationship. The distance between adjacent observation area surfaces is equal to the detection step size.

[0116] By dynamically setting the detection step size and establishing a one-to-one correspondence between observation points and observation area surfaces, a coordinated optimization of the tunnel spatial structure and data acquisition density is achieved. The detection step size can be adjusted in real time based on dynamic parameters such as tunnel ventilation speed and gas outflow rate. This not only avoids oversampling (high ventilation, low outflow scenarios) or detection blind spots (low ventilation, high outflow scenarios) caused by a fixed step size, but also ensures that the distance between adjacent observation areas accurately matches the gas diffusion characteristic scale, providing spatial continuity for subsequent modeling.

[0117] The observation area surface is progressively divided along the axial direction of the tunnel using the detection step length. Each observation point serves as a logical identification point of the corresponding area surface. Through the geometric definition of the axial range and radial full-section coverage, the discrete detection data is dynamically bound to the three-dimensional topological structure of the tunnel, solving the problem of local concentration field distortion caused by the decoupling of data and physical space in traditional point cloud scanning. In addition, for abnormal working conditions such as tunnel deformation or obstacle obstruction, the elastic step compensation algorithm maintains the mathematical consistency of the "point-surface" mapping relationship while ensuring continuous spatial coverage by correcting the observation area surface range and recalculating the total number of steps in real time, providing a highly reliable spatial reference framework for the gas diffusion model. This step is strongly associated with the spatial topology through dynamic parameter driving, significantly improving the input signal-to-noise ratio of subsequent data clustering and model iteration.

[0118] In one embodiment, before setting the detection step length of the inspection robot, the method further includes:

[0119] Controlling the inspection robot to move in the coal mine tunnel to collect gas concentration and main airflow direction;

[0120] Keeping other variables constant, the partial derivative of the main wind flow direction is calculated to obtain the gas concentration change rate; the gas concentration change rate is used to represent the degree of change in gas concentration when moving along the main wind flow direction.

[0121] The gas concentration change rate is the ratio of the gas concentration change rate to the distance moved along the main wind flow direction.

[0122] Coal mine tunnels are special spaces. When airflow enters the mine, it is constrained and blocked by the tunnel walls, forming a continuous and stable flow that differs from the atmospheric flow on the ground. The flow of air in coal mine tunnels can generally be considered a steady flow, with the airflow direction being along the tunnel's central axis. To study the variation in gas concentration along the main airflow direction, the partial derivative of the main airflow direction is taken while keeping other variables constant, yielding the rate of change of gas concentration along the main airflow direction. When the gas concentration change rate is greater than 0, it indicates that the gas concentration is increasing; when the gas concentration change rate is less than 0, it indicates that the gas concentration is decreasing; and when the gas concentration change rate approaches 0, it indicates that the gas concentration is not changing and is tending to be stable.

[0123] As an example, keeping other parameters unchanged, taking gas as an example, the change of gas concentration along the main air flow direction is as follows: Figure 2a-2l shown.

[0124] Depend on Figure 2a It can be seen that within a range of 1000 meters, the most rapid changes in gas are concentrated within 600 meters, the changes are more prominent within 200 meters, the changes gradually slow down from 200 meters to 400 meters, and there is basically no change from 400 meters to 600 meters. After 600 meters, the gas concentration tends to be stable and almost unchanged. In order to better study the specific changes of gas in the roadway, the roadway is segmented. See Figure 2b Figure 2l .

[0125] Figure 2b Indicates the rate of change of gas concentration along the wind direction within 200m. Figure 2c Indicates the rate of change of gas concentration along the wind direction within 10m, Figure 2d Indicates the rate of change of gas concentration along the wind direction within 10m to 20m. Figure 2e Indicates the rate of change of gas concentration along the wind direction within 20m to 30m. Figure 2f Indicates the rate of change of gas concentration along the wind direction within 30m to 40m. Figure 2g Indicates the rate of change of gas concentration along the wind direction within 40m to 100m. Figure 2hIndicates the rate of change of gas concentration along the wind direction within 100m to 200m. Figure 2i Indicates the rate of change of gas concentration along the wind direction within 300m to 400m. Figure 2j Indicates the rate of change of gas concentration along the wind direction within 400m to 500m. Figure 2k Indicates the rate of change of gas concentration along the wind direction within 500m to 600m. Figure 2l Indicates the rate of change of gas concentration along the wind direction within 600m to 700m.

[0126] according to Figure 2a to Figure 2l As shown, the gas concentration change rate r along the main airflow direction x is summarized as follows:

[0127] When x ≥ 340 m, |r| ≤ 1 × 10 -4 , when Δx=100m, Δc=|r|×x≤0.01%;

[0128] When 80m≤x≤340m, |r|≤1×10 -3 , when Δx=10m, Δc=|r|×x≤0.01%;

[0129] When 20m≤x≤80m, 1×10 -3 ≤|r|≤1×10 -4 , when Δx=10m, 0.01%≤Δc≤0.065%;

[0130] When x≤20m, close to the working surface, the change of |r| is more complicated.

[0131] In one embodiment, setting the detection step of the inspection robot includes:

[0132] The detection step length of the inspection robot is determined according to the gas concentration change rate; within the detection step length, the fluctuation of the gas concentration change rate is less than a threshold value.

[0133] In this invention, the inspection robot's detection step size is inversely proportional to the gas concentration change rate. When the gas concentration change rate is relatively large, indicating significant changes in gas concentration within this section of the roadway, the inspection robot's detection step size should be reduced, increasing the frequency of segmented modeling of coal mine roadway gas distribution to enable timely and accurate identification of hazardous areas. When the gas concentration change rate is relatively small, indicating minimal changes in gas concentration within this section of the roadway, the inspection robot's detection step size can be appropriately extended.

[0134] By taking the partial derivative with respect to the main airflow direction x, we can determine the rate of change r of gas concentration with distance from the main airflow direction x. Within a certain distance from the main airflow direction x, the rate of change r of gas concentration remains almost constant and can be considered a constant. In this case, the space within this distance from the main airflow direction x can be considered the inspection step of the inspection robot; the inspection time within this space is the ratio of the inspection step to the inspection robot's speed.

[0135] S102: Control the inspection robot to move along the lane, and the inspection robot collects gas environment data during the movement.

[0136] The gas environment data includes the tunnel size, wind speed and various gas concentrations collected by the inspection robot at the current detection point.

[0137] In the present invention, the inspection robot is equipped with a positioning system and a gas environment detection system. The positioning system can detect the inspection robot's position coordinates in the coal mine tunnel, the distance between the inspection robot and the tunnel wall, and the distance between the inspection robot and the tunnel roof. The gas environment detection system includes various sensors, such as methane sensors, carbon monoxide sensors, and temperature sensors.

[0138] In the present invention, the methane sensor, carbon monoxide sensor and temperature sensor are all suspended vertically, no more than 300 mm from the top plate and no less than 200 mm from the side wall (wall) of the tunnel. Figure 3 The inspection robot’s observation point, observation area, detection step length, and sensor installation location are shown in Figure 2. Figure 4 shown.

[0139] exist Figure 4 In, P i Indicates the point where the inspection robot collects gas environment information in the tunnel, i.e., the observation point. i It indicates the location area where methane and other sensors are installed as stipulated in the Coal Mine Safety Regulations, i.e., the observation area. For the inspection robot, the sensor installation position is fixed, and the gas concentration is at a constant height Z = Z R The measurement is done at the time of observation. Since the principle of inspection robot path planning is to avoid obstacles and ensure passage safety, the observation point P i As the inspection robot's walking trajectory changes, it is not necessarily on a straight line or a broken line, that is, the Y coordinate changes with the inspection robot's path.

[0140] When the inspection robot performs gas environment detection, when the inspection robot's driving speed is constant, the distance between two adjacent observation points is the detection step length, the detection time interval T is the same, and the distance between two adjacent observation area surfaces is the detection step length.

[0141] S103: The target gas environment data collected during the process of the inspection robot moving from the (i-1)th observation area surface to the (i)th observation area surface is used as a target gas environment data dataset.

[0142] The target gas environment data includes the tunnel size, wind speed and concentrations of various gases collected by the inspection robot at the current detection point.

[0143] The target observation area Ai represents the roadway height z∈(Z0-300,300], y∈(-(Y0-200),(Y0-200)), and the target observation point P(x,y,z)∈A i According to the Coal Mine Safety Regulations, any point on the target observation area can represent the gas concentration value of the section. Based on this, in the tunnel with the detection space step, the gas concentration value on a line formed by point P along the x-direction can also represent the gas concentration value of this tunnel space. In order to obtain a more certain unit tunnel gas concentration value, the DPC clustering algorithm is used to perform cluster analysis on the gas concentration values on a line formed by point P along the x-direction, and the peak with the largest density at this time is taken as the gas concentration value of this unit tunnel section.

[0144] S104: Calculate the target gas environment data set using a DPC clustering algorithm and output the i-1th cluster center.

[0145] To improve the inspection efficiency of patrol robots, it is necessary to extract valid data from large amounts of detection data, eliminate redundant data, and improve the data quality of coal mine roadway gas diffusion modeling. A clustering algorithm based on rapid search to find density peaks has the advantages of high computational efficiency, automatic acquisition of cluster numbers, applicability to irregularly shaped clusters, and insensitivity to noise. This paper improves the density peak clustering algorithm (DPC clustering algorithm) by eliminating the impact of noise on collected data, rapidly classifying and processing the data, and outputting the classification results, thus reducing the computational effort required by patrol robots to establish roadway gas diffusion models.

[0146] In one embodiment, the DPC clustering algorithm relative distance calculation method adopts the Mahalanobis distance calculation method.

[0147] The DPC clustering algorithm uses Euclidean distance to calculate relative distances. Traditional Euclidean distance, which represents the distance between two points in m-dimensional space, is significantly affected by differences in dimensions and orders of magnitude. This makes it difficult to accurately output cluster centers when processing complex datasets. Mahalanobis distance, on the other hand, represents the covariance distance of data and primarily considers correlations between data. It is unaffected by differences in dimensions and orders of magnitude, meeting the processing requirements of inspection robot data in this paper. To address these issues, this paper uses Mahalanobis distance instead of the Euclidean distance used in the original DPC clustering algorithm to calculate the cluster distance matrix.

[0148] The data collected by the coal mine roadway inspection robot's gas environment monitoring system includes information such as gas concentration, temperature and humidity, wind speed, current roadway space dimensions, and the inspection robot's coordinate position. This involves multiple dimensions and orders of magnitude. To eliminate this effect, the Mahalanobis distance is used instead of the Euclidean distance of the original DPC algorithm to improve the distance matrix. In the original DPC clustering algorithm, the determination of cluster centers requires manual selection. To avoid the increased clustering uncertainty caused by manually assigning cluster centers due to unclear cluster boundaries of the coal mine roadway inspection robot's gas environment monitoring data, a cluster center determination parameter is introduced to automatically select cluster centers.

[0149] In one embodiment, the Mahalanobis distance calculation method is:

[0150] Calculate the Mahalanobis distance between each target gas environment data in the target gas environment data set and the target gas environment data set:

[0151]

[0152] Wherein, μ is the overall mean of the target gas environment data in the target gas environment data set; V is the covariance matrix of the target gas environment data set, and x is the target gas environment data;

[0153] The covariance matrix is:

[0154]

[0155] Wherein, n is the total number of target gas environment data in the target gas environment data set, m means that each target gas environment data is a row vector containing m variables; v ij Represents the covariance of any two samples Xi and Xj in the same data set;

[0156] v ij =Cov(X i ,X j ),i,j=1,2,...n;

[0157] In the target gas environment data set, the Mahalanobis distance between the xth target gas environment data and the yth target gas environment data is:

[0158]

[0159] Wherein, x and y are two different target gas environment data in the target gas environment data set.

[0160] In one embodiment, the DPC clustering algorithm is used to calculate the target gas environment data set and output the i-1th cluster center, including:

[0161] Calculate the cluster center determination parameter of each target gas environment data in the target gas environment data set:

[0162] γ i =P i ·Δ i ;

[0163] Among them, γ i is the cluster center determination parameter of the i-th target gas environment data, P i is the local density ρ of the i-th target gas environment data i The normalized value, Δ i is the relative high density point distance δ to the i-th target gas environment data i Normalized value;

[0164]

[0165] Among them, ρ min is ρ i The minimum value in ρ max is ρ i The maximum value in ;

[0166]

[0167] Among them, δ max is δ i The minimum value in δ min is δ i The maximum value in ;

[0168] The target gas environment data with the largest cluster center determination parameter in the target gas environment data set is used as the target gas environment data corresponding to the i-1th cluster center.

[0169] According to the definition of clustering algorithms, the larger the γ value of an observation point, the greater the probability that it is the cluster center. When processing gas environment detection data from patrol robots, a threshold can be set to automatically select density peaks. The threshold is selected using a heuristic threshold method. Values greater than the threshold are determined as density peak points, i.e., the cluster center of the target gas environment dataset. The environmental parameters corresponding to this cluster center are the input parameters for optimizing the parameters of the coal mine roadway gas diffusion model.

[0170] S105: Using the target gas environment data corresponding to the i-1th cluster center as input parameters for optimizing parameters of a coal mine roadway gas diffusion model, and constructing a current space initial gas diffusion model based on a particle swarm optimization algorithm.

[0171] S106: Obtaining a calculated gas concentration value corresponding to the (i-1)th cluster center using the current spatial initial gas diffusion model.

[0172] S107: If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center meets a preset condition, the gas distribution of the i-1th sampling space is generated using the current space initial gas diffusion model.

[0173] The i-1th sampling space is the sampling space between the i-1th observation area surface and the i-th observation area surface.

[0174] S108: If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center does not meet a preset condition, reconstructing the current spatial initial gas diffusion model.

[0175] Until the difference between the calculated gas concentration value output by the initial gas diffusion model in the current space and the corresponding gas concentration sampling value meets the preset conditions.

[0176] S109: After obtaining the gas distribution of each sampling space, a coal mine roadway gas distribution map is generated according to the gas distribution of each sampling space.

[0177] This paper analyzes the concentration gradient of coal mine roadway gases along the wind flow direction and proposes setting the spatial step size of gas environment monitoring by coal mine roadway inspection robots based on the gas concentration gradient. This approach reduces the spatial monitoring step size in roadway sections with high gas concentration variation rates, and allows for more frequent modeling of coal mine roadway gas diffusion, enabling timely and accurate identification of hazardous areas.

[0178] Furthermore, the coal mine roadway inspection robot monitors relevant gas environment information along a given displacement path, employing an improved DPC clustering algorithm to process the gas environment data. The output cluster center pairs use this gas environment data as training and verification data for the gas diffusion model. This data is then input into the coal mine roadway gas diffusion model to optimize the gas diffusion coefficient and calculate the gas distribution within the detection space. As the inspection robot moves, gas concentration distribution data for different coal mine roadway sections along its displacement path is obtained. The calculated concentration of the previous detection space is verified using the gas concentration at the next detection space path point, and the roadway gas distribution model is continuously revised to better match the roadway environment, achieving the goal of using the roadway gas distribution model to calculate and generate the roadway gas distribution.

[0179] In order to better illustrate the feasibility of the inspection method provided by the present invention, the present invention conducts an experimental comparative explanation based on the above steps.

[0180] Figure 5 As shown, Figure 5The present invention provides a schematic diagram of the structure of a simple inspection robot. The simple inspection robot has a compact structure, flexible movement, and is convenient for conducting gas environment inspections in simulated tunnels. At the same time, in order to verify the accuracy of the gas environment detection results of the inspection robot, the present invention installs a movable bracket at the observation point in the simulated tunnel, such as Figure 6 As shown, the brackets on both sides (bracket 1 and bracket 2 2) can move in the vertical direction, and the middle bracket (bracket 3 3) can move in the horizontal direction. The sensor 4 (taking the chlorine sensor as an example) is installed on bracket 1 1, bracket 2 2, and bracket 3 3.

[0181] The inspection robot's lower computer transmits collected gas environment data to its upper computer via Ethernet. The upper computer runs a gas environment information monitoring and safety assessment platform system. This system primarily includes modules for real-time data acquisition, real-time data monitoring and display, evaluation result display, detection trajectory, data storage, and data playback.

[0182] In the gas environment information monitoring and safety evaluation operation platform system, the communication configuration and data acquisition module receives real-time data transmitted by the lower computer by setting the TCP / IP communication interface parameters, and displays it in the real-time data monitoring model. The display part is equipped with a dial display and a table display. The dial displays the instantaneous gas environment information value of the current detection trajectory coordinate position, and the table displays the information value near the current time. The regional information represents the gas environment information of the area. The evaluation result module displays the gas environment information safety level of the area, which is convenient for staff to make early warnings or maintenance decisions in time; the data storage module is mainly used to classify the real-time monitoring values and store them in a targeted manner for subsequent analysis and use; the data playback module is used to view the historical detection counts, which can be queried according to the tester's name and test conditions; the detection trajectory module can reflect the detection position coordinates in real time, which is convenient for monitoring personnel to understand the real-time monitoring position.

[0183] The airflow control system maintained the average wind speed in the simulated roadway at Vwind = 1.75 m / s. The gas flow control system controlled the helium outflow rate (Q) to 9.6 m³ / h. The inspection robot, at its initial position in the roadway, was 6 m from the working face (i.e., x = 6 m) and conducted inspections along the simulated return air roadway. The coal mine roadway inspection robot maintained a speed of approximately Vrobot = 0.2 m / s, and the coal mine roadway inspection robot's gas environment monitoring system had a detection frequency of 10 Hz.

[0184] To verify the reliability of the coal mine roadway inspection robot's gas environment inspection strategy, three detection lines were set up along the return air direction of the simulated roadway. These lines were labeled as detection line T1 (corresponding to bracket 1), detection line T2 (corresponding to bracket 2), and detection line T3 (corresponding to bracket 3). Detection lines T1 and T2 were installed on either side of the roadway at a height of 0.4m, 0.4m and 0.5m from the roadway's central axis, respectively. Detection line T3 was installed on the central axis of the roadway roof. Each detection line contained five helium sensors, arranged starting at 2.3m from the air outlet, with a 0.4m spacing between sensors on each detection line.

[0185] The inspection robot drove along the central axis of the simulated tunnel and collected about 1,400 data points. Some experimental gas environment data are shown in Table 1. The original data curve is shown in the figure below. Figure 7 shown.

[0186] Table 1 Experimental gas environment data (partial)

[0187]

[0188] From Table 1 experimental gas environment data (part) and Figure 7 The graph of the raw gas detection data shows that noise is present in the collected data due to the influence of other objective factors, such as the experimental environment. Furthermore, within a certain distance traveled by the inspection robot, the detection data appears to overlap and does not change. This may be due to the lack of change in the gas concentration within this area, or it may be due to the limited detection accuracy of the sensor, which cannot detect even subtle changes in the gas. To obtain valid data for modeling and eliminate redundant data, an improved DPC clustering algorithm was used to perform cluster analysis on the data in this detection space, quickly searching for the density peak cluster points of the gas detection dataset.

[0189] The clustering results are as follows Figure 8 As shown by Figure 8 It can be seen that the cluster centers after data processing are scattered in the gas environment dataset, effectively eliminating noise and redundant data. The output processing results of some improved DPC algorithms are shown in Table 2.

[0190] Table 2 Improved DPC algorithm output processing test results (partial)

[0191]

[0192] Using the cluster centers output by the improved DPC clustering algorithm (Table 2) as input and validation data, a particle swarm optimization (PSO) algorithm was used to optimize the model parameters and establish an initial gas diffusion model for the current space. The swarm particles were initialized to Size = 100, the particle dimension D = 5 (i.e., p1, p2, p3, p4, β), and the maximum number of iterations Tmax = 500. The learning factors C1 = C2 = 2, and a linearly decreasing dynamic inertia weight was used, with Wstart = 0.9 and Wend = 0.4.

[0193] like Figure 9 In the fitness function diagram shown, when the number of iterations reaches about 50, the fitness value reaches a stable level at 0.0015. Figure 10 In the shown gas concentration measured values and model calculated values, the curve is the model calculated value, and the points represent the actual detection values. When there is an error in the detection value, the model calculated value can basically fit the detection value.

[0194] As shown in Table 3, the results of the PSO algorithm for optimizing model parameters show that the correlation coefficient between the detection value and the model calculation value reaches 0.9640, with a high degree of fit. The mean square error of the model is 0.0387, indicating that the optimized coal mine roadway gas diffusion model is relatively effective.

[0195] Table 3 Number of detection points and model optimization parameters

[0196]

[0197] To further demonstrate the adaptability of the optimized roadway gas diffusion model and verify the feasibility of the gas environment inspection strategy for the coal mine roadway inspection robot, the gas concentration values on three detection lines were calculated based on this coal mine roadway gas diffusion model, labeled as calculated values C1, calculated values C2, and calculated values C3. These correspond to the sensor detection values at the corresponding coordinate positions of the three detection lines (detection line T1, detection line T2, and detection line T3), as shown in Table 4.

[0198] Table 4 Model calculated values and test line measured values

[0199]

[0200] In order to more intuitively observe the calculated values of the model and the actual measurement of the detection line, draw the following Figure 11 The comparison chart of calculated value and measured value is shown in the figure. Figure 11It can be seen that the calculated value of the coal mine tunnel gas diffusion model is basically consistent with the gas distribution trend of the corresponding detection line. The maximum absolute error of the gas concentration error of the two is 0.05%, the minimum absolute error is 0.03%, and the average absolute error is 0.04%, all of which are less than 5% of the detection value, meeting the requirements of the "Coal Mine Safety Regulations" on gas detection error, and can reflect the gas distribution law of coal mine tunnels, thereby proving that the gas environment inspection strategy of the coal mine tunnel inspection robot proposed in this invention is reliable and feasible.

[0201] In the invention, the target observation point M is located in the target observation area that is no more than 300mm away from the roof and no less than 200mm away from the side wall (wall) of the tunnel. Taking the center point of the target observation area as the target observation point, let the tunnel section height be H, the width be L, and the inspection robot travel distance in the return air direction be x, then the coordinates of the target observation point M are (x, H-150, 0). Inputting the coordinates of the target observation point M into the coal mine tunnel gas diffusion model established above, a gas concentration distribution line composed of the target observation point M can be obtained. Figure 12 shown.

[0202] According to safety assessment principles, the highest concentration of toxic and hazardous gases in a given area is considered a representative value. By establishing safety assessment indicators and models, and calculating the safety membership, the safety of the gas environment in that area is evaluated. Based on this, the gas concentration in the target observation area of the simulated coal mine tunnel is 1.76%.

[0203] Based on the above, the present invention further provides a tunnel gas environment inspection device based on a coal mine inspection robot, which is applied to the inspection robot performing inspections in coal mine tunnels. The device includes:

[0204] A setting module is used to set the detection step length of the inspection robot and determine i observation points and i observation area surfaces based on the lane length and the detection step length; wherein the observation points and observation area surfaces have a one-to-one correspondence, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; and the distance between the i-1-th observation area surface and the i-th observation area surface is the detection step length;

[0205] an acquisition module, configured to control the inspection robot to move along the lane, and collect gas environment data during the movement of the inspection robot; and to use the target gas environment data collected during the movement of the inspection robot from the i-1th observation area surface to the i-th observation area surface as the target gas environment data dataset;

[0206] A calculation module, configured to calculate the target gas environment data set using a DPC clustering algorithm and output an i-1th cluster center;

[0207] A construction module is used to use the target gas environment data corresponding to the i-1th cluster center as input parameters for optimizing the parameters of the coal mine tunnel gas diffusion model, and to construct an initial gas diffusion model of the current space based on a particle swarm optimization algorithm;

[0208] An analysis module, configured to obtain a calculated gas concentration value corresponding to the i-1th cluster center using a current spatial initial gas diffusion model;

[0209] An output module, configured to generate the gas distribution of the i-1 sampling space using the initial gas diffusion model of the current space if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center meets a preset condition; the i-1 sampling space is the sampling space between the i-1 observation area surface and the i-th observation area surface; if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center does not meet the preset condition, reconstruct the initial gas diffusion model of the current space until the difference between the calculated gas concentration value output by the initial gas diffusion model of the current space and the corresponding sampled gas concentration value meets the preset condition;

[0210] The generation module is used to generate a coal mine roadway gas distribution map according to the gas distribution of each sampling space after obtaining the gas distribution of each sampling space.

[0211] In one embodiment, the DPC clustering algorithm relative distance calculation method adopts the Mahalanobis distance calculation method.

[0212] In one embodiment, the Mahalanobis distance calculation method is:

[0213] Calculate the Mahalanobis distance between each target gas environment data in the target gas environment data set and the target gas environment data set:

[0214]

[0215] Wherein, μ is the overall mean of the target gas environment data in the target gas environment data set; V is the covariance matrix of the target gas environment data set, and x is the target gas environment data;

[0216] The covariance matrix is:

[0217]

[0218] Wherein, n is the total number of target gas environment data in the target gas environment data set, m means that each target gas environment data is a row vector containing m variables; v ij Represents the covariance of any two samples Xi and Xj in the same data set;

[0219] v ij =Cov(X i ,X j ),i,j=1,2,...n;

[0220] In the target gas environment data set, the Mahalanobis distance between the xth target gas environment data and the yth target gas environment data is:

[0221]

[0222] Wherein, x and y are two different target gas environment data in the target gas environment data set.

[0223] In one embodiment, the building block is specifically configured to:

[0224] Calculate the cluster center determination parameter of each target gas environment data in the target gas environment data set:

[0225] γ i =P i ·Δ i ;

[0226] Among them, γ i is the cluster center determination parameter of the i-th target gas environment data, P i is the local density ρ of the i-th target gas environment data i The normalized value, Δ i is the relative high density point distance δ to the i-th target gas environment data i Normalized value;

[0227]

[0228] Among them, ρ min is ρ i The minimum value in ρ max is ρ i The maximum value in ;

[0229]

[0230] Among them, δ max is δ i The minimum value in δ min is δ i The maximum value in ;

[0231] The target gas environment data with the largest cluster center determination parameter in the target gas environment data set is used as the target gas environment data corresponding to the i-1th cluster center.

[0232] In one embodiment, the acquisition module is further configured to:

[0233] Controlling the inspection robot to move in the coal mine tunnel to collect gas concentration and main airflow direction;

[0234] Keeping other variables constant, the partial derivative of the main wind flow direction is calculated to obtain the gas concentration change rate; the gas concentration change rate is used to represent the degree of change in gas concentration when moving along the main wind flow direction.

[0235] In one embodiment, the setting module is specifically configured to:

[0236] The detection step length of the inspection robot is determined according to the gas concentration change rate; within the detection step length, the fluctuation of the gas concentration change rate is less than a threshold value.

[0237] In one embodiment, the target gas environment data includes the tunnel size, wind speed, and concentrations of multiple gases collected by the inspection robot at the current detection point.

[0238] In addition, an embodiment of the present application provides a control device, including a processor and a memory, wherein the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to complete the tunnel gas environment inspection method based on the coal mine inspection robot.

[0239] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which is loaded by a processor to execute the tunnel gas environment inspection method based on a coal mine inspection robot.

[0240] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The above-mentioned program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0241] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. The replacement may be a replacement of a portion of a structure, device, or method step, or it may be a complete technical solution. Any equivalent replacement or modification based on the technical solution and inventive concept of the present invention shall be covered by the scope of protection of the present invention.

Claims

1. A tunnel gas environment inspection method based on a coal mine inspection robot is characterized by: The method comprises: The inspection step length of the inspection robot is set, and i observation points and i observation area surfaces are determined according to the lane length and the inspection step length; wherein the observation points and the observation area surfaces correspond one to one, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; and the distance between the i-1-th observation area surface and the i-th observation area surface is the inspection step length; Controlling the inspection robot to move along the lane, and collecting gas environment data during the movement of the inspection robot; The target gas environment data collected during the inspection robot's movement from the (i-1)th observation area to the (i)th observation area is used as a target gas environment data set; The DPC clustering algorithm is used to calculate the target gas environment data set and output the i-1th cluster center; The target gas environment data corresponding to the i-1th cluster center is used as the input parameter for optimizing the parameters of the coal mine tunnel gas diffusion model, and the current space initial gas diffusion model is constructed based on the particle swarm optimization algorithm; Using the current spatial initial gas diffusion model, a calculated gas concentration value corresponding to the i-1th cluster center is obtained; If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center meets the preset conditions, the gas distribution of the i-1th sampling space is generated using the current spatial initial gas diffusion model; the i-1th sampling space is the sampling space between the i-1th observation area surface and the i-th observation area surface; If the difference between the calculated gas concentration value corresponding to the i-1th cluster center and the sampled gas concentration value corresponding to the i-1th cluster center does not meet the preset condition, the current spatial initial gas diffusion model is rebuilt until the difference between the calculated gas concentration value output by the current spatial initial gas diffusion model and the corresponding sampled gas concentration value meets the preset condition; After obtaining the gas distribution of each sampling space, a coal mine roadway gas distribution map is generated according to the gas distribution of each sampling space.

2. The method according to claim 1, characterized in that The relative distance calculation method of the DPC clustering algorithm adopts the Mahalanobis distance calculation method.

3. The method according to claim 2, characterized in that The Mahalanobis distance calculation method is: Calculate the Mahalanobis distance between each target gas environment data in the target gas environment data set and the target gas environment data set: Wherein, μ is the overall mean of the target gas environment data in the target gas environment data set; V is the covariance matrix of the target gas environment data set, and x is the target gas environment data; The covariance matrix is: Wherein, n is the total number of target gas environment data in the target gas environment data set, and m means that each target gas environment data is a row vector containing m variables; because j =Cov(X i ,Y j ),i,j=1,2,...n; In the target gas environment data set, the Mahalanobis distance between the xth target gas environment data and the yth target gas environment data is: Wherein, x and y are two different target gas environment data in the target gas environment data set.

4. The method according to claim 1, wherein The DPC clustering algorithm is used to calculate the target gas environment data set and output the i-1th cluster center, including: Calculate the cluster center determination parameter of each target gas environment data in the target gas environment data set: c i =P i ·D i ; Among them, γ i is the cluster center determination parameter of the i-th target gas environment data, P i is the local density ρ of the i-th target gas environment data i The normalized value, Δ i is the relative high density point distance δ to the i-th target gas environment data i Normalized value; Among them, ρ min is ρ i The minimum value in ρ max is ρ i The maximum value in ; Among them, δ max is δ i The minimum value in δ min is δ i The maximum value in ; The target gas environment data with the largest cluster center determination parameter in the target gas environment data set is used as the target gas environment data corresponding to the i-1th cluster center.

5. The method according to claim 1, wherein Before setting the detection step length of the inspection robot, the method further includes: Controlling the inspection robot to move in the coal mine tunnel to collect gas concentration and main airflow direction; Keeping other variables constant, the partial derivative of the main wind flow direction is calculated to obtain the gas concentration change rate; the gas concentration change rate is used to represent the degree of change in gas concentration when moving along the main wind flow direction.

6. The method according to claim 5, characterized in that The step of setting the inspection step of the inspection robot includes: The detection step length of the inspection robot is determined according to the gas concentration change rate; within the detection step length, the fluctuation of the gas concentration change rate is less than a threshold value.

7. The method according to claim 1, characterized in that The target gas environment data includes the tunnel size, wind speed and multiple gas concentrations collected by the inspection robot at the current detection point.

8. A tunnel gas environment inspection device based on a coal mine inspection robot, characterized in that: The device comprises: A setting module is used to set the detection step length of the inspection robot and determine i observation points and i observation area surfaces based on the lane length and the detection step length; wherein the observation points and observation area surfaces have a one-to-one correspondence, the i-th observation point is any point in the i-th observation area surface; i is an integer greater than 1; and the distance between the i-1-th observation area surface and the i-th observation area surface is the detection step length; an acquisition module, configured to control the inspection robot to move along the lane, and collect gas environment data during the movement of the inspection robot; and to use the target gas environment data collected during the movement of the inspection robot from the i-1th observation area surface to the i-th observation area surface as the target gas environment data dataset; A calculation module, configured to calculate the target gas environment data set using a DPC clustering algorithm and output an i-1th cluster center; A construction module is used to use the target gas environment data corresponding to the i-1th cluster center as input parameters for optimizing the parameters of the coal mine tunnel gas diffusion model, and to construct an initial gas diffusion model of the current space based on a particle swarm optimization algorithm; An analysis module, configured to obtain a calculated gas concentration value corresponding to the i-1th cluster center using a current spatial initial gas diffusion model; An output module, configured to generate the gas distribution of the i-1 sampling space using the initial gas diffusion model of the current space if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center meets a preset condition; the i-1 sampling space is the sampling space between the i-1 observation area surface and the i-th observation area surface; if the difference between the calculated gas concentration value corresponding to the i-1 cluster center and the sampled gas concentration value corresponding to the i-1 cluster center does not meet the preset condition, reconstruct the initial gas diffusion model of the current space until the difference between the calculated gas concentration value output by the initial gas diffusion model of the current space and the corresponding sampled gas concentration value meets the preset condition; The generation module is used to generate a coal mine roadway gas distribution map according to the gas distribution of each sampling space after obtaining the gas distribution of each sampling space.

9. A control device, characterized in that: It includes a processor and a memory, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to complete the tunnel gas environment inspection method based on the coal mine inspection robot as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that A computer program is stored, and the computer program is loaded by a processor to execute the tunnel gas environment inspection method based on a coal mine inspection robot as described in any one of claims 1 to 7.

Images