Method for positioning tiny leakage of gas extraction pipeline based on multi-sensor fusion

Abstract

The invention discloses a gas extraction pipeline tiny leakage positioning method based on multi-sensor fusion, which belongs to the technical field of positioning, is used for gas extraction pipeline leakage positioning, and comprises the following steps: establishing a state equation and an observation equation of a motion process of an inspection robot, and realizing inspection robot positioning under a navigation system based on Kalman filtering. And realizing coal mine gas extraction pipeline leakage source positioning under a microphone array coordinate system based on a beam forming algorithm, and obtaining coal mine gas extraction pipeline leakage source positioning under a navigation system through coordinate conversion. According to the method, the leakage source position of the gas extraction pipeline under the navigation system is directly given, maintainers can conveniently and quickly reach the leakage source position, and the first-aid repair time is greatly shortened; the positioning error of the inspection robot can be effectively reduced, and the phenomenon that the positioning error is accumulated along with time when the inspection robot only uses an accelerometer and a gyroscope for positioning is avoided; the influence of background noise on a positioning result can be greatly reduced, and the spatial resolution of a leakage positioning result is improved.

Description

Technical Field

[0001] The invention discloses a method for locating minute leakage of a gas extraction pipeline based on multi-sensor fusion, and belongs to the technical field of positioning. Background Art

[0002] Once a gas extraction pipeline leaks, the methane concentration and oxygen concentration in the pipeline will quickly reach the explosion limit range. At the same time, there are many possible ignition sources inside and outside the pipeline, which can easily cause a gas explosion. Therefore, timely positioning of the leak in the early stage of a gas extraction pipeline leak is of great significance to preventing major safety accidents caused by gas explosions and ensuring efficient gas extraction and utilization. The existing gas extraction pipeline leakage positioning methods include three methods: manual inspection, distributed monitoring, and robot inspection. Manual inspection refers to the inspection personnel holding a gas leak detection instrument to conduct close point-by-point inspection of the gas extraction pipeline. The detection efficiency of manual inspection is very low and seriously threatens the life safety of the inspection personnel. Distributed monitoring refers to the monitoring of gas leaks by arranging flow meters and pressure sensors at multiple points on the gas extraction pipeline, or arranging gas sensors, fiber optic vibration sensors, fiber optic temperature sensors and other types of sensors at multiple points in the return air lane. The positioning error of distributed monitoring is large and it cannot detect small leaks. Robot inspection refers to the detection of gas extraction pipeline leaks by carrying various gas, temperature and other types of sensors on the inspection robot. Most of the existing coal mine gas inspection robots are directly transplanted from industries such as gas pipeline inspection and high-voltage power grid inspection, and have not been deeply developed for on-site application scenarios in the return air channel. In essence, they are gas and temperature type sensors installed on a moving chassis, following the sensor technology in distributed monitoring. The problems existing in distributed monitoring have not been fundamentally solved, and there are still problems such as large leakage positioning errors and inability to locate tiny leakage sources. Summary of the invention

[0003] The purpose of the present invention is to provide a method for locating tiny leaks in gas extraction pipelines based on multi-sensor fusion, so as to solve the problems in the prior art that robot inspection has large leakage positioning errors and cannot locate the source of tiny leaks.

[0004] A method for locating tiny leaks in a gas extraction pipeline based on multi-sensor fusion includes establishing state equations and observation equations for the motion process of a patrol robot, realizing the positioning of the patrol robot in a navigation system based on Kalman filtering, realizing the positioning of the leakage source of a coal mine gas extraction pipeline in a microphone array coordinate system based on a beamforming algorithm, and obtaining the positioning of the leakage source of a coal mine gas extraction pipeline in a navigation system through coordinate transformation.

[0005] The state equations for establishing the inspection robot's motion process include:

[0006]

[0007]

[0008] Where P x,k , P y,k , P z,k They represent the position coordinates in the x, y, and z directions at time k, respectively, and v x,k 、v y,k 、v z,k They represent the velocity components in the x, y, and t directions at time k, respectively. s is the sampling interval, f s is the sampling frequency, a nx 、a ny 、a nz represents the acceleration components in the x, y, and z directions in the navigation system, a bx 、a by 、a bz It represents the acceleration components in the x, y, and z directions under the load system, which are obtained by the accelerometer carried by the inspection robot.

[0009] [a bx 、a by 、a bz ] T After the rotation matrix transformation, we get [a nx 、a ny 、a nz ] T :

[0010]

[0011] Where g is the acceleration due to gravity, It represents the attitude conversion matrix, ψ is the heading angle, θ is the pitch angle, γ is the roll angle, and the northeast celestial coordinate system is adopted. The system is located at the upper right front, and the heading angle is positive by north-west. ψ, θ, and γ are obtained by the gyroscope carried by the inspection robot.

[0012] The state equation of the inspection robot's motion process is written in matrix form:

[0013] X k =FX k-1 +Bu k-1 +w k-1 ;

[0014] X k =[P x,k P y,k P z,k v x,k v y,k v z,k ] T ;

[0015]

[0016] u k-1 =[a nx,k-1 a ny,k-1 a nz,k-1 ] T ;

[0017] In the formula, X k is the coordinate velocity matrix, F is the state transfer matrix, B is the input control matrix, u k-1 is the system input, w k-1 represents the system process noise, w k-1 The covariance matrix of is the system process noise covariance matrix Q.

[0018] The observation equations for the motion process of the inspection robot include the movement of the inspection robot with unknown position coordinates (x, y, z) in the navigation system to the i-th inspection robot at the known position (x i ,y i ,z i )’s UWB base station measurement distance is r i , 1≤i≤m, i is the base station number, m is the total number of base stations:

[0019]

[0020] Use a i represents the coordinate vector of the i-th base station, and Z represents the coordinate vector of the inspection robot:

[0021] a i =(x i ,y i ,z i );

[0022] Z=(x,y,z) T .

[0023] The observation equations for the motion process of the inspection robot are established, and the matrix form of the measured distance is:

[0024] AZ = b;

[0025]

[0026] In the formula, A is the coefficient matrix, b is the measurement distance matrix, and the least squares solution of the ultra-wideband base station to the inspection robot position coordinates is:

[0027] Z=(A T A) -1 A T b.

[0028] The observation equation for the motion process of the inspection robot is established by converting the position coordinate Z of the inspection robot at time k calculated by the ultra-wideband base station intok As observation, the observation equation of the system is obtained according to the least squares solution:

[0029] Z k =HX k +n k ;

[0030]

[0031] Where H is the system observation matrix, n k Represents the system observation noise, and the corresponding observation noise covariance matrix is ​​R.

[0032] The positioning of the inspection robot in the navigation system based on Kalman filtering includes:

[0033] S1. Determine F and B according to the matrix form of the state equation of the inspection robot's motion process, and determine the position coordinate Z of the inspection robot at time k. k Determine H;

[0034] S2, obtain Q and R;

[0035] S3, given the initial state vector at time k = 1 and the error covariance matrix is a non-zero matrix, and the superscript ^ indicates an estimated value;

[0036] S4, state vector at time k = 2, 3, ..., N and the error covariance matrix Obtained through the following iterative formula of Kalman filtering:

[0037]

[0038] In the formula, the upper right subscript - indicates the prior, K k represents the Kalman gain at time k;

[0039] S5, state vector The first three items P x,k , P y,k , P z,k is the position of the inspection robot in the x, y, and z directions in the navigation system at time k. The position of the inspection robot in the navigation system is

[0040]

[0041] The leakage source location of coal mine gas extraction pipeline in the microphone array coordinate system based on the beamforming algorithm includes: when the sound signal collected by the microphone is at the highest sound pressure level SPL in the ultrasonic frequency band max When the threshold η is exceeded, it is considered that the gas extraction pipeline has leaked;

[0042] The microphone array carried by the inspection robot consists of M microphones, where m = 1, ..., M represents the microphone index, and the highest sound pressure level SPL in the ultrasonic frequency band is recorded. max The corresponding frequency is f max , the time domain sound pressure signal collected by the microphone array is divided into L time blocks, the overlap between blocks is set to 50%, and the frequency f is obtained by discrete Fourier transform of the time domain signal max The frequency domain signal under is p, and the frequency f is obtained max The cross-spectral matrix C under is:

[0043]

[0044] In the formula, the upper right subscript * indicates that the conjugate transpose of the matrix is ​​required;

[0045] The pipeline acoustic imaging area is divided into N grid points, each of which is a potential leakage source, n = 1, ..., N represents the grid point index, represents the position coordinates of the nth grid point in the pipeline acoustic imaging area in the microphone array coordinate system. The acoustic power matrix B of the pipeline acoustic imaging area is obtained by the beamforming algorithm:

[0046]

[0047] In the formula, C NoDiag is the matrix obtained by setting the main diagonal elements of the cross-spectrum matrix C to 0, W is the acoustic transfer function matrix between the microphone array and the pipeline acoustic imaging area, and the acoustic transfer function matrix W is calculated using the acoustic finite element method;

[0048] The acoustic imaging area grid point corresponding to the maximum value in the acoustic power matrix B The location of the leakage source in the coal mine gas extraction pipeline in the microphone array coordinate system as the leakage source location for:

[0049]

[0050] according to as well as Obtain the location of the leakage source of the coal mine gas extraction pipeline in the navigation system through coordinate transformation for:

[0051]

[0052] In the formula, R SG represents the transformation matrix from the carrier system to the navigation system, R SG Equivalent to R ASrepresents the transformation matrix from the microphone array coordinate system to the carrier system, ΔX represents the position coordinate of the origin of the microphone array in the carrier system, and R AS and ΔX both represent the installation error of the microphone array on the inspection robot, which is obtained from the installation calibration error.

[0053] Compared with the prior art, the present invention has the following beneficial effects: by fusing the measurement data of four sensors, namely, accelerometer, gyroscope, UWB and microphone array, the leakage source position of the gas extraction pipeline under the navigation system can be directly given, which is convenient for maintenance personnel to quickly reach the leakage source position and greatly shorten the repair time; the Kalman filter algorithm is used to fuse the accelerometer, gyroscope and UWB data, which can effectively reduce the positioning error of the inspection robot and avoid the phenomenon that the positioning error accumulates over time when the inspection robot only uses accelerometer and gyroscope for positioning; the conventional beamforming algorithm has very strong robustness, and at the same time, combined with the method of removing the main diagonal line of the microphone array measurement signal cross spectrum matrix, it can greatly reduce the influence of background noise on the positioning result and improve the spatial resolution of the leakage positioning result; the acoustic transfer function matrix between the inspection robot and the gas extraction pipeline is calculated by the acoustic finite element method, which not only avoids the disadvantage that the free field Green's function cannot consider the multipath propagation effect of sound waves, but also makes full use of the characteristic that the relative position of the inspection robot and the gas extraction pipeline in the return air channel remains unchanged, and there is no need to perform multiple calculations due to the change of the position of the inspection robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The present invention is a technical flow chart of the method for locating gas extraction pipeline leakage. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] A method for locating tiny leaks in a gas extraction pipeline based on multi-sensor fusion includes establishing state equations and observation equations for the motion process of a patrol robot, realizing the positioning of the patrol robot in a navigation system based on Kalman filtering, realizing the positioning of the leakage source of a coal mine gas extraction pipeline in a microphone array coordinate system based on a beamforming algorithm, and obtaining the positioning of the leakage source of a coal mine gas extraction pipeline in a navigation system through coordinate transformation.

[0057] The state equations for establishing the inspection robot's motion process include:

[0058]

[0059] Where P x,k , P y,k , P z,k They represent the position coordinates in the x, y, and z directions at time k, respectively, and v x,k 、v y,k 、v z,k They represent the velocity components in the x, y, and z directions at time k, respectively, and t s is the sampling interval, f s is the sampling frequency, a nx 、a ny 、a nz represents the acceleration components in the x, y, and z directions in the navigation system, a bx 、a by 、a bz It represents the acceleration components in the x, y, and z directions under the load system, which are obtained by the accelerometer carried by the inspection robot.

[0060] [a bx 、a by 、a bz ] T After the rotation matrix transformation, we get [a nx 、a ny 、a nz ] T :

[0061]

[0062] Where g is the acceleration due to gravity, It represents the attitude conversion matrix, ψ is the heading angle, θ is the pitch angle, γ is the roll angle, and the northeast celestial coordinate system is adopted. The system is located at the upper right front, and the heading angle is positive by north-west. ψ, θ, and γ are obtained by the gyroscope carried by the inspection robot.

[0063] The state equation of the inspection robot's motion process is written in matrix form:

[0064] X k =FX k-1 +Bu k-1 +w k-1 ;

[0065] X k =[P x,k P y,k P z,k v x,k v y,k v z,k ] T ;

[0066]

[0067] uk-1 =[a nx,k-1 a ny,k-1 a nz,k-1 ] T ;

[0068] In the formula, X k is the coordinate velocity matrix, F is the state transfer matrix, B is the input control matrix, u k-1 is the system input, w k-1 represents the system process noise, w k-1 The covariance matrix of is the system process noise covariance matrix Q.

[0069] The observation equations for the motion process of the inspection robot include the movement of the inspection robot with unknown position coordinates (x, y, z) in the navigation system to the i-th inspection robot at the known position (x i ,y i ,z i )’s UWB base station measurement distance is r i , 1≤i≤m, i is the base station number, m is the total number of base stations:

[0070]

[0071] Use a i represents the coordinate vector of the i-th base station, and Z represents the coordinate vector of the inspection robot:

[0072] a i =(x i ,y i ,z i );

[0073] Z=(x,y,z) T .

[0074] The observation equations for the motion process of the inspection robot are established, and the matrix form of the measured distance is:

[0075] AZ = b;

[0076]

[0077] In the formula, A is the coefficient matrix, b is the measurement distance matrix, and the least squares solution of the ultra-wideband base station to the inspection robot position coordinates is:

[0078] Z=(A T A) -1 A T b.

[0079] The observation equation for the motion process of the inspection robot is established by converting the position coordinate Z of the inspection robot at time k calculated by the ultra-wideband base station into kAs observation, the observation equation of the system is obtained according to the least squares solution:

[0080] Z k =HX k +n k ;

[0081]

[0082] Where H is the system observation matrix, n k Represents the system observation noise, and the corresponding observation noise covariance matrix is ​​R.

[0083] The positioning of the inspection robot in the navigation system based on Kalman filtering includes:

[0084] S1. Determine F and B according to the matrix form of the state equation of the inspection robot's motion process, and determine the position coordinate Z of the inspection robot at time k. k Determine H;

[0085] S2, obtain Q and R;

[0086] S3, given the initial state vector at time k = 1 and the error covariance matrix is a non-zero matrix, and the superscript ^ indicates an estimated value;

[0087] S4, state vector at time k = 2, 3, ..., N and the error covariance matrix Obtained through the following iterative formula of Kalman filtering:

[0088]

[0089] In the formula, the upper right subscript - indicates the prior, K k represents the Kalman gain at time k;

[0090] S5, state vector The first three items P x,k , P y,k , P z,k is the position of the inspection robot in the x, y, and z directions in the navigation system at time k. The position of the inspection robot in the navigation system is

[0091]

[0092] The leakage source location of coal mine gas extraction pipeline in the microphone array coordinate system based on the beamforming algorithm includes: when the sound signal collected by the microphone is at the highest sound pressure level SPL in the ultrasonic frequency band max When the threshold η is exceeded, it is considered that the gas extraction pipeline has leaked;

[0093] The microphone array carried by the inspection robot consists of M microphones, where m = 1, ..., M represents the microphone index, and the highest sound pressure level SPL in the ultrasonic frequency band is recorded. max The corresponding frequency is f max , the time domain sound pressure signal collected by the microphone array is divided into L time blocks, the overlap between blocks is set to 50%, and the frequency f is obtained by discrete Fourier transform of the time domain signal max The frequency domain signal under is p, and the frequency f is obtained max The cross-spectral matrix C under is:

[0094]

[0095] In the formula, the upper right subscript * indicates that the conjugate transpose of the matrix is ​​required;

[0096] The pipeline acoustic imaging area is divided into N grid points, each of which is a potential leakage source, n = 1, ..., N represents the grid point index, represents the position coordinates of the nth grid point in the pipeline acoustic imaging area in the microphone array coordinate system. The acoustic power matrix B of the pipeline acoustic imaging area is obtained by the beamforming algorithm:

[0097]

[0098] In the formula, C NoDiag is the matrix obtained by setting the main diagonal elements of the cross-spectrum matrix C to 0, W is the acoustic transfer function matrix between the microphone array and the pipeline acoustic imaging area, and the acoustic transfer function matrix W is calculated using the acoustic finite element method;

[0099] The acoustic imaging area grid point corresponding to the maximum value in the acoustic power matrix B The location of the leakage source in the coal mine gas extraction pipeline in the microphone array coordinate system as the leakage source location for:

[0100]

[0101] according to as well as Obtain the location of the leakage source of the coal mine gas extraction pipeline in the navigation system through coordinate transformation for:

[0102]

[0103] In the formula, R SG represents the transformation matrix from the carrier system to the navigation system, R SG Equivalent to R AS represents the transformation matrix from the microphone array coordinate system to the carrier system, ΔX represents the position coordinate of the origin of the microphone array in the carrier system, and RAS and ΔX both represent the installation error of the microphone array on the inspection robot, which is obtained from the installation calibration error.

[0104] The present invention proposes a gas extraction pipeline leakage positioning method based on multi-sensor data fusion, installs an ultra-wideband (UWB) base station on the wall of the return air channel, installs an accelerometer, a gyroscope, a UWB tag, and a microphone array on the inspection robot, and uses the characteristics of ultrasound generated when a small leak occurs in the gas extraction pipeline to achieve high-precision positioning of small leaks in the gas extraction pipeline through multi-sensor data fusion. The coordinate system where the inspection robot is located is the carrier system, and the coordinate system where the UWB base station is located is the navigation system. The accelerometer can measure the acceleration of the inspection robot, the gyroscope can measure the angular velocity of the inspection robot, the UWB tag can obtain the distance from each UWB base station, and the microphone array can collect leakage sound signals.

[0105] In the present invention, since the background noise is mainly concentrated on the main diagonal of the cross-spectrum matrix, setting the main diagonal elements of the cross-spectrum matrix C to 0 can greatly reduce the impact of background noise on the positioning results and improve the spatial resolution of the leakage positioning results. Since pipeline leakage may occur on the back of the pipeline, the leakage sound signal received by the microphone array at this time includes both leakage sound that penetrates the pipeline and reflected sound reflected by the wall and diffracted sound diffracted by the pipeline surface. Therefore, the free-field Green's function cannot be directly used to establish the sound transfer function matrix W. Based on the characteristic that the relative positions of the inspection robot and the gas extraction pipeline in the return air channel remain unchanged, the present invention uses the acoustic finite element method to calculate the sound transfer function matrix W. Only one offline calculation is required to apply the sound transfer function matrix at all inspection positions, which fully considers the multipath propagation effect of the leakage sound in the return air channel, and does not require multiple calculations due to changes in the position of the inspection robot.

[0106] The present invention proposes a method for locating the leakage source of a gas extraction pipeline by fusing the measurement data of four sensors, namely, accelerometer, gyroscope, UWB and microphone array, in a navigation system. First, the position of the inspection robot in the navigation system is obtained by using the accelerometer, gyroscope and UWB, and then the position of the leakage source in the carrier system is obtained by using the microphone array. Finally, the position of the leakage source of the gas extraction pipeline in the navigation system is obtained by the coordinate conversion from the carrier system to the navigation system. Based on the characteristic that the relative positions of the inspection robot and the gas extraction pipeline in the return air lane remain unchanged, a method for calculating the acoustic transfer function matrix between the inspection robot and the gas extraction pipeline by using the acoustic finite element method is proposed.

[0107] Technology process integration Figure 1As shown in the figure, the state equation of the patrol robot's motion process is established through the acceleration and posture of the patrol robot. At the same time, the position of the patrol robot is obtained through the UWB base station and UWB tag as the observation equation of the patrol robot's motion process. The Kalman filter algorithm is used to fuse the measurement data of the accelerometer, gyroscope, and UWB to realize the position solution of the patrol robot in the navigation system. Since ultrasonic signals will appear at the leakage source position, the leakage ultrasonic signals are received by the microphone array and post-processed using the conventional beamforming algorithm to realize the leakage positioning in the microphone array coordinate system. Finally, the leakage position of the gas extraction pipeline in the navigation system is obtained through coordinate transformation.

[0108] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some or all of the technical features may be replaced by equivalents, and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for locating tiny leaks in gas extraction pipelines based on multi-sensor fusion, characterized in that: It includes establishing the state equation and observation equation of the inspection robot's motion process, realizing the positioning of the inspection robot in the navigation system based on Kalman filtering, realizing the positioning of the coal mine gas extraction pipeline leakage source in the microphone array coordinate system based on the beamforming algorithm, and obtaining the positioning of the coal mine gas extraction pipeline leakage source in the navigation system through coordinate transformation.

2. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 1 is characterized in that: The state equations for establishing the inspection robot's motion process include: Where P x,k , P y,k , P z,x They represent the position coordinates in the x, y, and z directions at time k, respectively, and v x,k 、v y,k 、v z,k They represent the velocity components in the x, y, and z directions at time k, respectively, and t s is the sampling interval, f s is the sampling frequency, a nx 、a ny 、a nz represents the acceleration components in the x, y, and z directions in the navigation system, a bx 、a by 、a bz It represents the acceleration components in the x, y, and z directions under the load system, which are obtained by the accelerometer carried by the inspection robot.

3. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 2 is characterized in that: [a bx 、a by 、a bz ] T After the rotation matrix transformation, we get [a nx 、a ny 、a nz ] T : Where g is the acceleration due to gravity, It represents the attitude conversion matrix, ψ is the heading angle, θ is the pitch angle, γ is the roll angle, and the northeast celestial coordinate system is adopted. The system is located at the upper right front, and the heading angle is positive by north-west. ψ, θ, and γ are obtained by the gyroscope carried by the inspection robot.

4. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 3 is characterized in that: The state equation of the inspection robot's motion process is written in matrix form: X k =FX k-1 +This k-1 +w k-1 ; X k =[P x,k P y,k P z,k v x,k v y,k v z,k ] T ; u k-1 =[a nx,k-1 a ny,k-1 a nz,k-1 ] T ; Where, X k is the coordinate velocity matrix, F is the state transfer matrix, B is the input control matrix, u k-1 is the system input, w k-1 represents the system process noise, w k-1 The covariance matrix of is the system process noise covariance matrix Q.

5. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 4 is characterized in that: The observation equations for the motion process of the inspection robot include the movement of the inspection robot with unknown position coordinates (x, y, z) in the navigation system to the i-th inspection robot at the known position (x i ,y i ,z i )’s UWB base station measurement distance is r i , 1≤i≤m, i is the base station number, m is the total number of base stations: Use a i represents the coordinate vector of the i-th base station, and Z represents the coordinate vector of the inspection robot: a i =(x i ,y i ,z i ); Z=(x,y,z) T 。 6. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 5 is characterized in that: The observation equations for the motion process of the inspection robot are established, and the matrix form of the measured distance is: AZ = b; In the formula, A is the coefficient matrix, b is the measurement distance matrix, and the least squares solution of the ultra-wideband base station to the inspection robot position coordinates is: Z=(A T A) -1 A T b。 7. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 6 is characterized in that: The observation equation for the motion process of the inspection robot is established by converting the position coordinate Z of the inspection robot at time k calculated by the ultra-wideband base station into k As observation, the observation equation of the system is obtained according to the least squares solution: Z k =HX k +n k ; Where H is the system observation matrix, n k Represents the system observation noise, and the corresponding observation noise covariance matrix is ​​R.

8. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 7 is characterized in that: The positioning of the inspection robot in the navigation system based on Kalman filtering includes: S1. Determine F and B according to the matrix form of the state equation of the inspection robot's motion process, and determine the position coordinate Z of the inspection robot at time k. k Determine H; S2, obtain Q and R; S3, given the initial state vector at time k = 1 and the error covariance matrix is a non-zero matrix, and the superscript ^ indicates an estimated value; S4, state vector at time k = 2, 3, ..., N and the error covariance matrix Obtained through the following iterative formula of Kalman filtering: In the formula, the upper right subscript - represents the prior, and Kk represents the Kalman gain at time k; S5, state vector The first three items P x,k , P y,k , P z,k is the position of the inspection robot in the x, y, and z directions in the navigation system at time k. The position of the inspection robot in the navigation system is 9. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 8, characterized in that: The leakage source location of coal mine gas extraction pipeline in the microphone array coordinate system based on the beamforming algorithm includes: when the sound signal collected by the microphone is at the highest sound pressure level SPL in the ultrasonic frequency band max When the threshold η is exceeded, it is considered that the gas extraction pipeline has leaked; The microphone array carried by the inspection robot consists of M microphones, where m = 1, ..., M represents the microphone index, and the highest sound pressure level SPL in the ultrasonic frequency band is recorded. max The corresponding frequency is f max , the time domain sound pressure signal collected by the microphone array is divided into L time blocks, the overlap between blocks is set to 50%, and the frequency f is obtained by discrete Fourier transform of the time domain signal max The frequency domain signal under is p, and the frequency f is obtained max The cross-spectral matrix C under is: In the formula, the upper right subscript * indicates that the conjugate transpose of the matrix is ​​required; The pipeline acoustic imaging area is divided into N grid points, each of which is a potential leakage source, n = 1, ..., N represents the grid point index, represents the position coordinates of the nth grid point in the pipeline acoustic imaging area in the microphone array coordinate system. The acoustic power matrix B of the pipeline acoustic imaging area is obtained by the beamforming algorithm: In the formula, C NoDiag is the matrix obtained by setting the main diagonal elements of the cross-spectrum matrix C to 0, W is the acoustic transfer function matrix between the microphone array and the pipeline acoustic imaging area, and the acoustic transfer function matrix W is calculated using the acoustic finite element method; The acoustic imaging area grid point corresponding to the maximum value in the acoustic power matrix B The location of the leakage source in the coal mine gas extraction pipeline in the microphone array coordinate system as the leakage source location for:

10. The method for locating micro-leakage in a gas extraction pipeline based on multi-sensor fusion according to claim 9, characterized in that: according to as well as Obtain the location of the leakage source of the coal mine gas extraction pipeline in the navigation system through coordinate transformation for: In the formula, R SG represents the transformation matrix from the carrier system to the navigation system, R SG Equivalent to R AS represents the transformation matrix from the microphone array coordinate system to the carrier system, ΔX represents the position coordinate of the origin of the microphone array in the carrier system, and R AS and ΔX both represent the installation error of the microphone array on the inspection robot, which is obtained from the installation calibration error.

Images