Method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectrum technology

Abstract

The invention discloses a method for remotely measuring gas concentration two-dimensional distribution based on an absorption spectrum technology, and belongs to the technical field of gas monitoring and remote measurement. On the basis that the unmanned aerial vehicle is used for mounting a telemetering laser receiving and transmitting device to inspect the natural gas pipeline along the line, as the natural gas pipeline is buried underground, when a leakage point appears on the natural gas pipeline, a leaked natural gas mass is formed on the ground surface where the natural gas pipeline is buried; a remote measuring laser receiving and transmitting device carried by the unmanned aerial vehicle is used for detecting an area where the natural gas concentration exceeds a threshold value, and therefore the natural gas pipeline leakage point position is judged. According to the method, the high natural gas area can be more accurately positioned, so that the pipeline leakage point is deduced, the local constraint variable strategy is adopted, the resolving stability is improved, and artifacts or invalid data are avoided. And real-time generation and updating of a concentration distribution diagram are realized through simultaneous scanning and resolving.

Description

Technical Field

[0001] The invention relates to a method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectrum technology, belonging to the technical field of gas monitoring and remote measurement. Background Art

[0002] In recent years, absorption spectroscopy technology has been widely used in the field of gas detection due to its high sensitivity and accuracy, especially tunable semiconductor laser absorption spectroscopy (TDLAS) technology. However, the existing TDLAS technology is mainly limited to fixed monitoring points, making it difficult to achieve comprehensive monitoring over a large range. The rapid development of drone technology has provided a new solution for gas monitoring. By integrating TDLAS sensors into drones, dynamic inspections of natural gas pipelines can be achieved, and two-dimensional distribution data of gas concentration can be effectively obtained.

[0003] TDLAS technology can be combined with computed tomography (CT) technology, and reconstruction using projection or iterative methods can achieve two-dimensional (2D) temperature distribution and component concentration distribution measurement in the measurement area. This technology is mostly based on the fixed position of the laser and the detector, which surrounds a scanning plane with multiple parallel beams scanning at multiple angles of 0°-180° in the plane. For example: the combustion field plane is surrounded by multiple fixed lasers and detectors, and CT-TDLAS is used to reconstruct the temperature and concentration of this plane slice. Under this condition, the row vectors of the projection matrix L formed by the full-angle parallel beam scanning have low repeatability, and the formed projection matrix L usually has high stability. The solution of the reconstruction vector α can be directly solved by ART algebraic iteration.

[0004] However, CT-TDLAS technology encounters certain problems when combined with telemetry to reconstruct gas concentration distribution. The problems are mainly reflected in the following two points:

[0005] 1. CT-TDLAS is mostly based on multiple fixed lasers and detectors, and solves the concentration information in the scanning plane surrounded by them; while the telemetry process uses a laser transceiver, in which the laser and detector are integrated and located on the same side, and changes in position and angle with the drone and gimbal during the inspection process.

[0006] 2. Due to the limitation of the reflection plane during the inspection process, the detectors on one side cannot achieve full-angle scanning in the traditional CT-TDLAS process. In the scanning process with limited angles, the stability of the formed projection matrix L will have certain problems, making the reconstruction equation L*α=A an ill-posed equation. The solution of the reconstruction vector α is unstable, which affects the concentration reconstruction result in the concentration reconstruction area. Summary of the invention

[0007] In view of the shortcomings of the existing technology, a method for remotely measuring the two-dimensional distribution of gas concentration based on absorption spectroscopy technology is provided. The steps are simple, the area of ​​natural gas leakage can be obtained according to the laser detection data, and the location of the leakage point can be accurately determined.

[0008] To achieve the above technical objectives, the present invention discloses a method for remotely measuring the two-dimensional distribution of gas concentration based on absorption spectroscopy technology, which uses a remote sensing laser transceiver mounted on an unmanned aerial vehicle to inspect the natural gas pipeline along the line. Since the natural gas pipeline is buried underground, when a leak occurs in the natural gas pipeline, a leaked natural gas mass will be formed on the surface where the natural gas pipeline is buried. The remote sensing laser transceiver carried by the unmanned aerial vehicle is used to detect the area where the natural gas concentration exceeds the threshold, thereby determining the location of the leak in the natural gas pipeline.

[0009] The specific steps are as follows:

[0010] S1. Set the scanning route of the drone according to the laying conditions of the natural gas pipeline to be inspected. The detection route of the drone needs to cover the laying area of ​​the natural gas pipeline. The drone starts to cruise horizontally along the laying route of the natural gas pipeline at a height of 8 meters directly above the horizontal buried natural gas pipeline. From any section, the horizontal cruising route of the drone and the natural gas pipeline are simplified into two parallel lines in a two-dimensional plane. These two parallel lines are used to form a detection plane between the X-axis and the Z-axis. The detection plane is divided into multiple non-overlapping 2m×2m discrete grids. Every 16 consecutive discrete grids are combined into a concentration reconstruction area R of 8m×8m.

[0011] S2. On the detection plane, a drone scanning site is set every 2 meters according to the discrete grid size. The drone needs to hover every time it flies to a scanning site, and use the telemetry laser transceiver to emit laser scanning at 9 different angles in the current concentration reconstruction area R. The surface natural gas content information is obtained by emitting laser signals using absorption spectroscopy technology. According to the position of the drone in the concentration reconstruction area R, 5 laser data are selected from the 9 laser data of each scanning site to ensure that the drone obtains 25 non-overlapping laser information that can completely cover the current concentration reconstruction area R at each scanning site;

[0012] S3. 25 laser beams that completely cover a concentration reconstruction area R are received by a telemetry laser transceiver after being reflected by the ground. The telemetry laser transceiver obtains the light intensity information after being absorbed by natural gas and subtracts the background fitting to obtain the absorbance of natural gas. The absorbance reflects the integrated concentration (ppm*m) of natural gas on the beam path. The effective path length information L of each light in the 25 lasers is recorded. m,j And its corresponding integrated concentration information A m ;

[0013] S4, constructing multiple continuous and partially overlapping concentration reconstruction region sets R'{R1, R2...R i}, where the concentration reconstruction region set R'{R1, R2...R i The area of ​​the concentration reconstruction area R2 in} is an 8m×8m area that is constructed after the drone moves 2m. The first 12 discrete grids of the concentration reconstruction area R2 overlap with the last 12 discrete grids of R1 until R at the end of the detection plane. i ;

[0014] S5. Use the ART algebraic iterative reconstruction method to solve the gas concentration α in 16 discrete grids in the first 8m×8m concentration reconstruction area R1 j , j = 1, 2...16; then use the last 12 discrete grid solutions known in R1 to calculate the solution of R2, so as to determine that all 16 discrete grids in R2 have stable solutions, and use the ART iteration method to solve the concentration reconstruction concentration region set R'{R1, R2...R i}, complete the reconstruction and update of concentration from local to global stability; obtain the specific concentration distribution in each discrete grid in two-dimensional space

[0015] S6. The natural gas concentration of each detected discrete grid is marked, and then a two-dimensional distribution map of the natural gas concentration of the natural gas pipeline on the ground in the drone detection route is drawn based on the natural gas concentration information of all the marked discrete grids, and updated in real time. It can accurately locate the area with high natural gas, not only can the location of the pipeline leakage point be accurately fed back for rapid maintenance, but also the personnel near the leakage point can be quickly evacuated according to the generated two-dimensional distribution map of natural gas concentration to prevent poisoning and other disasters.

[0016] Furthermore, each time the UAV flies to a scanning location, it uses a telemetry laser transceiver to emit laser scanning at 9 different angles. In this process, the 9 angle scanning is controlled by the gimbal of the telemetry laser transceiver. The telemetry laser transceiver is connected to the UAV through the gimbal. The scanning angles are θ0 as the median perpendicular line, θ0 as 0°, θ1 as -45° to the left of the median perpendicular line θ0, and θ2 as -45° to the left of the median perpendicular line θ1. -4 , -36.9° angle θ -3 , -26.6° angle θ -2 , -14.0° angle θ -1 , and θ1 with an angle of 14.0°, θ2 with an angle of 26.6°, θ3 with an angle of 36.9°, and θ4 with an angle of 45° to the right of the perpendicular bisector θ0;

[0017] The process of selecting 5 laser beam data from the laser rays generated at 9 angles is as follows: the scanning beam data at five angles θ0, θ1, θ2, θ3, and θ4 are used at the first position in the current concentration reconstruction area R, and the scanning beam data at five angles θ0, θ1, θ2, θ3, and θ4 are used at the second position. -1 , θ0, θ1, θ2, θ3 scanning beam data, the third position uses θ -2 ,θ -1 , θ0, θ1, θ2 scanning beam data, the fourth point θ -3 ,θ -2 ,θ -1 , θ0, θ1 scanning beam data, the fifth position θ -4 ,θ -3 ,θ -2 ,θ -1 , θ0 scanning beam data, so as to ensure that t When the concentration is within the range, the UAV has 5 valid beam data at each scanning position, and a total of 25 non-overlapping laser information that can completely cover the concentration reconstruction area R is obtained, completely covering the concentration reconstruction area R.

[0018] Furthermore, the integrated concentrations corresponding to the 25 laser beams are transmitted to the data processing platform via wireless communication to form 25 data. The 25 data are used to form a 25-dimensional projection column vector A (ppm*m). The length of the j-th grid that the i-th beam passes through is recorded as L i，j , by L i，j Construct a 25*16-dimensional projection matrix L(m), and define the natural gas concentration inside the j-th grid as α j , α1-α 16 The 16-dimensional reconstruction vector α is the concentration reconstruction result. The natural gas concentrations of the 16 discrete grids are represented by α1-α 16 express;

[0019] Known:

[0020]

[0021] In short:

[0022] L·α=A

[0023] Where A is the projection result vector composed of the real-time updated path integral concentration information obtained during the inspection process, and L is the projection matrix known from the optical path layout. Solving the reconstruction vector α is to obtain the concentration distribution information of the natural gas in the inspection plane.

[0024] Furthermore, the effective path length L in each discrete grid is i,j Determined by the width w of the discrete grid and the angle θ between the light and the coordinate axis; the integral concentration A corresponding to each beam path i Depend on Measured, where I0 and I t are the incident light intensity and the transmitted light intensity, P is the pressure, S(T) is the line intensity, L i,j To absorb the effective path length, the pressure and line intensity in the entire measurement area are set to constant values ​​at normal temperature and pressure.

[0025] Furthermore, the equation L·α=A is solved by ART algebraic iteration to calculate the gas concentration α in 16 discrete grids in the first 8m*8m concentration reconstruction area R1. j , the ART algebraic iteration process is as follows:

[0026] The initial gas concentration distribution is vector α (0) Under the premise of (0) =0, the projection iteration formula for reconstructing the vector α is as follows:

[0027]

[0028] Where i is the number of iterations, α (i) is the concentration reconstruction result of the i-th iteration, A i is the projection result of the i-th beam, that is, the integrated concentration of natural gas on the i-th beam path, L i is the projection coefficient of the i-th light ray, that is, the i-th row vector of the matrix L, λ is the relaxation parameter, λ=0.001.

[0029] (1) Assign values ​​to the gas concentrations in the 16 discrete grids, and assign the initial value of the natural gas concentration of all discrete grids in the concentration reconstruction area R1 to 0. (i) =0, i=0,j=1,2…,16;

[0030] (2) For the ith laser ray among the 25 lasers, the estimated projection result A on the ith row of the equation is calculated using the following formula: i * , that is, solving the estimated integrated concentration of natural gas on the i-th laser path:

[0031]

[0032] (3) The error Δ between the actual projection and the estimated projection on the i-th row of equations is calculated using the following formula: i , thus obtaining the error between the actual integrated natural gas concentration on the ith laser path and the estimated integrated natural gas concentration on the ith laser path:

[0033]

[0034] (4) Calculate the corrected value of the gas concentration in the jth discrete grid:

[0035]

[0036] (5) Use the correction value to correct the gas concentration value in the jth discrete grid:

[0037]

[0038] (6) The result of the previous round is used as the initial value, and the process of (2) to (5) is repeated until the convergence condition is reached; each round of complete iteration is 25 times. If the number of iterations i>25, the next round of iteration is carried out. Each round of iteration uses the result of the previous round of iteration as the initial value, that is, the initial value of the next round of iteration α( 0 ) is taken as the value of the previous round of α( 25 ) until the preset convergence condition is met, and the convergence condition meets the maximum number of iterations or the error convergence criterion:

[0039] ||a (i) ||-||a (i-1) ||≤ε(e=10 -5 )

[0040] The natural gas concentration distribution of 16 discrete grids in R1 is calculated as follows: α1-α 16 , concentration reconstruction concentration region set R'{R1, R2...R i The remaining elements in} are obtained by repeating the contents of (1)-(6).

[0041] Furthermore, the natural gas concentrations α1-α 16 , as a reference value to constrain some free variables in the concentration reconstruction region R2:

[0042] The ART algorithm is combined with telemetry, and the strategy of local constrained variables is adopted. Part of the data solved last time is used as a reference. α1, α2, α3, and α5 in R2 are selected as the four redundant free variables to constrain. Since α1, α2, α3, and α5 in R2 and α5, α6, α7, and α9 in R1 are the same discrete grid, and the values ​​of α1, α2, α3, and α5 in R2 are equal to those of α5, α6, α7, and α9 in R1, α5, α6, α7, and α9 in R1 are used to solve the remaining 12 unknown solutions in the re-formed R2. The four free variables α1, α2, α3, and α5 in the projection equation L*α=A to be solved in R2 are constrained by α5, α6, α7, and α9 solved by R1, that is, α1, α2, α3, and α5 in R2 are replaced by α5, α6, α7, and α9 in R1. The resulting equation is:

[0043]

[0044] Converted into the following form:

[0045]

[0046] Simplified to:

[0047] L R2 ×α R2 =A R2

[0048] The only unknown numbers left are α4, α6, α7, α8, α9, α 10 , α 11 , α 12 , α 13 , α 14 , α 15 , α 16 , thus forming a new unknown number of 12, the projection matrix L R2 The equation to be solved is of rank 12, and other solution values ​​in R1 can be used as initial values ​​to participate in the ART algebraic iterative reconstruction process of the projection matrix in R2;

[0049] R t The concentration distribution inside is given by R t-1 The solved variable constraints within are constrained and ART algebraic iterative reconstruction is performed cyclically to complete the natural gas concentration of all concentration reconstruction areas R on the entire UAV route.

[0050] A drone comprises a drone body. A remote sensing laser transceiver for detecting gas concentration information by laser is arranged below the drone body through a gimbal. The type of gas detected can be adjusted according to actual needs. A computing device is arranged inside the drone. The computing device comprises a processor and a memory. The processor is electrically connected to the memory. The memory is used to store instructions and data. The processor is used to execute a method for remote sensing two-dimensional distribution of gas concentration based on decomposition and absorption spectroscopy technology.

[0051] Beneficial effects: This solution combines telemetry technology with the ART algorithm and applies it to the field of mobile inspection. By planning the scanning path, the integral concentration can be used to invert the natural gas distribution on the scanning plane, and the high natural gas area can be more accurately located to infer the pipeline leakage point. At the same time, when combining telemetry with the ART algorithm, improvements have been made to the ART algorithm, using lasers and detectors to scan the detection plane at all angles. The inversion results are relatively accurate, and more accurate and stable concentration distribution reconstruction can be achieved during the telemetry process. In mobile telemetry, due to the need to use the road surface where the buried pipeline is located as a reflection background, the laser and detector are mounted on one side, and the angle is limited. It is difficult for existing methods to generate a stable projection matrix. Therefore, when using the ART algorithm, the strategy of local constrained variables is adopted to improve the solution stability and avoid the appearance of artifacts or invalid data. And by solving while scanning, the concentration distribution map can be generated and updated in real time.

[0052] The advantages of the present invention are as follows: This method has good local stability: limiting the free variables in each initial solution helps to control the stability of the local solution of the underdetermined system. This method has enhanced iterative convergence, and using partially stable solutions to constrain the free variables to iteratively solve the problem so that the reconstruction process gradually approaches global convergence, which helps to obtain an accurate global solution. This method has strong adaptability: it is suitable for systems with underdetermined properties, so that a reasonable concentration distribution solution can be obtained even under a rank-deficient projection matrix. Through this step-by-step introduction, local-to-global ART iterative solution scheme, a more accurate and stable concentration distribution reconstruction can be achieved during the telemetry process. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic flow chart of a method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology of the present invention;

[0054] Figure 2 This is a schematic diagram of the composition of a telemetry laser transceiver device used in a drone in an embodiment of the present invention;

[0055] Figure 3 This is a schematic diagram of the principle of using a telemetry laser transceiver device in a drone according to an embodiment of the present invention;

[0056] Figure 4 This is a schematic diagram of the structure of a telemetry laser transceiver used in a drone according to an embodiment of the present invention;

[0057] Figure 5 A schematic diagram of a discrete grid formed by using laser detection concentration reconstruction area in an embodiment of the present invention;

[0058] Figure 6 Schematic diagram of nine laser scans performed by a drone after hovering in an embodiment of the present invention;

[0059] Figure 7This is a schematic diagram of the remote sensing laser optical path layout of the concentration reconstruction area in an embodiment of the present invention;

[0060] Figure 8 Schematic diagram of the optical paths of the concentration reconstruction areas R1 and R2 in an embodiment of the present invention. Specific implementation plan

[0061] The embodiments of the present invention are further described below in conjunction with the accompanying drawings:

[0062] like Figure 1 As shown, the present invention discloses a method for remotely measuring the two-dimensional distribution of gas concentration based on absorption spectroscopy technology, which uses a drone carrying a laser transceiver to inspect the natural gas pipeline, and uses the ART algebraic iteration method of cyclic constraint variables to invert the distribution of natural gas concentration in the detection plane, and draw a two-dimensional distribution cloud map of natural gas concentration in the grid space. If the natural gas concentration is higher than the threshold, the abnormal concentration position is scanned and reconstructed twice, and the abnormal concentration area after verification is located and an alarm is issued, so as to realize rapid and comprehensive inspection and leakage location of the natural gas pipeline.

[0063] To solve the concentration distribution in the entire 8m*S m rectangular area. The specific performance is:

[0064] The specific steps are as follows:

[0065] S1: The drone scanning route is set according to the laying conditions of the natural gas pipeline to be inspected. The drone inspection route needs to cover the laying area of ​​the natural gas pipeline. The drone starts horizontal cruising along the laying route of the natural gas pipeline at a height of 8 meters above the horizontal buried natural gas pipeline. From any section, the drone horizontal cruising route and the natural gas pipeline are simplified into two parallel lines in a two-dimensional plane. These two parallel lines are used to form a detection plane between the X-axis and the Z-axis. The detection plane is divided into 8m×8m concentration reconstruction areas R. Each concentration reconstruction area R includes 16 2m×2m discrete grids, such as Figure 5 As shown;

[0066] S2: The UAV sets a scanning point every 2 meters on the horizontal cruising route according to the discrete grid size. Every time the UAV flies to the scanning point, it needs to hover and use the telemetry laser transceiver to launch laser scanning at 9 different angles, and use the absorption spectrum technology to obtain the surface natural gas content information by emitting light signals. Select 5 laser data from the 9 laser data at each scanning point to ensure that the UAV obtains 25 non-overlapping laser information that can completely cover the concentration reconstruction area R at each scanning point, and completely covers each concentration reconstruction area R, such as Figure 6 As shown;

[0067] S3: 25 laser beams that completely cover a concentration reconstruction area R are received by the telemetry laser transceiver after being reflected by the ground. The detector obtains the light intensity information after absorption by natural gas and subtracts the background fitting to obtain the absorbance of natural gas. The absorbance reflects the integrated concentration of natural gas gas on the beam path (ppm*m). The effective path length information L of each light in the 25 lasers is recorded. m,j And its corresponding integrated concentration information A m ;

[0068] S4: The entire UAV detection route is decomposed into multiple continuous concentration reconstruction areas R using the ART iterative method and iteratively calculated. The gas concentration α in 16 discrete grids in the first 8m × 8m concentration reconstruction area R is solved by the ART algebraic iterative reconstruction method. j , j = 1, 2...16; then use the known solution in the first concentration reconstruction area R to solve the solution of the next area, so as to determine that the next area is a stable solution; in this process, it is continuously iterated, by solving the natural gas concentration information of each discrete grid, and continuously iterating, so as to obtain stable and accurate ground natural gas information;

[0069] S5: After the first concentration reconstruction area R1 is solved, the UAV continues to cruise and continues to solve the subsequent concentration reconstruction areas R i The ART algebraic iterative reconstruction of the constraint variables is performed cyclically to complete the reconstruction and update of the concentration from local to global stability. The drone first scans the 8m range to form the first 8m×8m concentration reconstruction area R1. Then, it scans 2m backward every t-1 times. The new scanned area overlaps with the last 6m*8m area of ​​the previous area to form the next 8m×8m concentration reconstruction area R t ,like Figure 7 and Figure 8 shown.

[0070] S6: The natural gas concentration of each detected discrete grid is marked, and then a two-dimensional distribution map of the natural gas concentration of the natural gas pipeline on the ground in the drone detection route is drawn based on the natural gas concentration information of all the marked discrete grids, and the map is updated in real time, and concentration thresholds are set to issue early warnings for areas with concentrations exceeding the limit.

[0071] The remote sensing laser transceiver device specifically solves the concentration X of the gas to be measured in each grid. j as follows:

[0072] Determine the scanning route of the drone based on the laying conditions of the natural gas pipeline to be inspected. The inspection route of the drone needs to cover the laying area of ​​the natural gas pipeline.

[0073] The UAV starts cruising 8m above the horizontal buried pipeline. The cruising route and the horizontal buried pipeline constitute a detection plane, which is divided into discrete grids. The concentration of each grid in the plane constitutes the concentration distribution in the plane. The grid size is 2m*2m. Assuming that the length of the buried pipeline to be detected is Sm, the detection plane is a rectangular area of ​​8m*Sm.

[0074] In order to reconstruct the concentration distribution of the 2m*2m discrete grid in the detection plane, the UAV sets a scanning site every 2m on the horizontal cruising route and performs multi-angle scanning.

[0075] Every time the UAV flies to the scanning position, it uses the telemetry laser transceiver to launch laser scanning at 9 different angles. In this process, the 9 angle scanning is controlled by the gimbal of the telemetry laser transceiver. The telemetry laser transceiver is connected to the UAV through the gimbal. The scanning angles are θ0 as the median perpendicular line, θ0 is 0°, and forms θ with the left side of the median perpendicular line θ0. -4 (-45°), θ -3 (-36.9°),θ -2 (-26.6°),θ -1 (-14.0°), and θ1(14.0°), θ2(26.6°), θ3(36.9°), and θ4(45°) are formed on the right side of the perpendicular bisector θ0; the fan angle setting makes it possible for the inspection point to be located at the scanning position (A,8) formed by the inspection plane with the starting point of the pipeline as the origin and 1m as the unit length, and to scan (A-8,0), (A-6,0), (A-4,0), (A-2,0), (A,0), (A+2,0), (A+4,0), (A+6,0), and (A+8,0).

[0076] Under this scanning optical path layout, the scanning position (A, 8) covers at most a square area of ​​8m*8m on both sides of the symmetry axis (A, 8) and (A, 0). Therefore, the minimum reconstructed solution area is a square area of ​​8m*8m, which contains 16 discrete grids. Assuming that the scan starts from the starting scanning position (0, 8), after the scan is completed at (8, 8), the first minimum concentration reconstruction area is completely covered. There are 25 scanning laser beams that completely cover this reconstruction area, and the remaining beams are partially used in the subsequent area reconstruction.

[0077] The 25 laser beams that completely cover the first concentration reconstruction area are received by the telemetry laser transceiver after reflection. The detector obtains the light intensity information after absorption by natural gas, and obtains the absorbance of natural gas after background subtraction and fitting. The absorbance reflects the integrated concentration (ppm*m) of natural gas gas on the beam path. These steps can be implemented by hardware processing circuits.

[0078] The integrated concentrations corresponding to the 25 lasers are transmitted to the data processing platform via wireless communication. These 25 data form a 25-dimensional projection column vector A (ppm*m).

[0079] The 16 discrete grids in the first reconstruction area completely cover 25 laser beams. The length of the jth grid that the mth beam passes through is recorded as L m，j , by L m，j Construct a 25*16 dimensional projection matrix L(m).

[0080] Define the natural gas concentration inside the jth grid as α j , α j It forms a 16-dimensional column vector α, where α is the reconstruction vector, that is, the concentration reconstruction result.

[0081] Known:

[0082]

[0083] In short:

[0084] L·α=A

[0085] Where A is the projection result vector composed of the real-time updated path integrated concentration information obtained during the inspection process, and L is the projection matrix known by the optical path layout. From the above, it can be seen that each time in the reconstruction area composed of a 4*4 grid, 25 beams completely cover the reconstruction area to form a 25*16-dimensional projection matrix L.

[0086] At this time, the problem is transformed into solving the reconstruction vector α composed of the concentration information in the discrete grid that reflects the concentration distribution information in the inspection plane when the projection matrix L and the projection result A are known.

[0087] like Figure 2 , Figure 3 and Figure 4 As shown, the laser transceiver includes a package shell, a collimating laser is provided at the end of the package shell, a detector and a data processing and communication module are provided inside the package shell, and the collimating laser includes a Fresnel lens and a collimating lens, and the collimating lens is connected to the laser and the driving power supply through an optical fiber. The laser transceiver is mounted on the drone through a pan-tilt platform, and the inspection route and detection plane are planned according to the buried position of the pipeline to realize horizontal cruising and fixed-point scanning.

[0088] The laser transceiver realizes the emission and reception of laser. The absorbance is calculated by receiving the light intensity information of each scanning beam, thereby inverting the integrated concentration on the beam path, and the above process is realized through the signal processing circuit. The integrated concentration information on each beam path is transmitted to the data processing platform through the wireless communication module.

[0089] The laser scanning device itself can only obtain the integrated concentration information (ppm*m) on a single optical path, but cannot directly obtain the specific concentration distribution within each discrete grid in the two-dimensional space. The ART algorithm iteratively reconstructs the scanning data through the projection matrix, infers the gas concentration of each discrete grid, and generates a high-resolution concentration distribution map. The integrated concentration data of a single optical path cannot identify the specific location of the concentration anomaly on the laser path. It reflects the integral amount of the concentration on the path, and it is difficult to determine the specific location of the natural gas leak. Natural gas leaks usually form a concentration gradient around the leak source. The ART algorithm can identify the area with the fastest concentration change and infer the dew point position through a high-resolution concentration distribution map.

[0090] The concentration distribution cloud map is drawn based on the calculated natural gas gas concentration distribution data and updated in real time. The concentration threshold is set and early warning is implemented for areas with concentration exceeding the limit.

Claims

1. A method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology, characterized in that: Based on the use of drones equipped with telemetry laser transceivers to inspect natural gas pipelines along the route, since natural gas pipelines are buried underground, when a leak occurs in the natural gas pipeline, a leaking natural gas cloud will be formed on the surface where the natural gas pipeline is buried. The remote sensing laser transceiver carried by the drone is used to detect areas where the natural gas concentration exceeds the threshold, thereby determining the location of the leak in the natural gas pipeline; The specific steps are as follows: S1. Set the scanning route of the drone according to the laying conditions of the natural gas pipeline to be inspected. The detection route of the drone needs to cover the laying area of ​​the natural gas pipeline. The drone starts to cruise horizontally along the laying route of the natural gas pipeline at a height of 8 meters directly above the horizontal buried natural gas pipeline. From any section, the horizontal cruising route of the drone and the natural gas pipeline are simplified into two parallel lines in a two-dimensional plane. These two parallel lines are used to form a detection plane between the X-axis and the Z-axis. The detection plane is divided into multiple non-overlapping 2m×2m discrete grids. Every 16 consecutive discrete grids are combined into a concentration reconstruction area R of 8m×8m. S2. On the detection plane, a drone scanning site is set every 2 meters according to the discrete grid size. The drone needs to hover every time it flies to a scanning site, and use the telemetry laser transceiver to emit laser scanning at 9 different angles in the current concentration reconstruction area R. The surface natural gas content information is obtained by emitting laser signals using absorption spectroscopy technology. According to the position of the drone in the concentration reconstruction area R, 5 laser data are selected from the 9 laser data of each scanning site to ensure that the drone obtains 25 non-overlapping laser information that can completely cover the current concentration reconstruction area R at each scanning site; S3. 25 laser beams that completely cover a concentration reconstruction area R are received by a telemetry laser transceiver after being reflected by the ground. The telemetry laser transceiver obtains the light intensity information after being absorbed by natural gas and subtracts the background fitting to obtain the absorbance of natural gas. The absorbance reflects the integrated concentration (ppm*m) of natural gas on the beam path. The effective path length information L of each light in the 25 lasers is recorded. m,j And its corresponding integrated concentration information A m ; S4, constructing multiple continuous and partially overlapping concentration reconstruction region sets R'{R1, R2...R i }, where the concentration reconstruction region set R'{R1, R2...R i The area of ​​the concentration reconstruction area R2 in} is an 8m×8m area that is constructed after the drone moves 2m. The first 12 discrete grids of the concentration reconstruction area R2 overlap with the last 12 discrete grids of R1 until R at the end of the detection plane. i ; S5. Use the ART algebraic iterative reconstruction method to solve the gas concentration α in 16 discrete grids in the first 8m×8m concentration reconstruction area R1 j , j = 1, 2...16; then use the last 12 discrete grid solutions known in R1 to calculate the solution of R2, so as to determine that all 16 discrete grids in R2 have stable solutions, and use the ART iteration method to solve the concentration reconstruction concentration region set R'{R1, R2...R i }, complete the reconstruction and update of concentration from local to global stability; obtain the specific concentration distribution in each discrete grid in two-dimensional space S6. The natural gas concentration of each detected discrete grid is marked, and then a two-dimensional distribution map of the natural gas concentration of the natural gas pipeline on the ground in the drone detection route is drawn based on the natural gas concentration information of all the marked discrete grids, and updated in real time. It can accurately locate the area with high natural gas, not only can the location of the pipeline leakage point be accurately fed back for rapid maintenance, but also the personnel near the leakage point can be quickly evacuated according to the generated two-dimensional distribution map of natural gas concentration to prevent poisoning and other disasters.

2. The method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology according to claim 1, characterized in that: Each time the UAV flies to the scanning position, it uses the telemetry laser transceiver to launch laser scanning at 9 different angles. In this process, the 9 angle scanning is controlled by the gimbal of the telemetry laser transceiver. The telemetry laser transceiver is connected to the UAV through the gimbal. The scanning angles are θ0 as the median perpendicular line, θ0 as 0°, θ1 as -45° to the left of the median perpendicular line θ0, and θ2 as -45° to the left of the median perpendicular line θ1. -4 , -36.9° angle θ -3 , -26.6° angle θ -2 , -14.0° angle θ -1 , and θ1 with an angle of 14.0°, θ2 with an angle of 26.6°, θ3 with an angle of 36.9°, and θ4 with an angle of 45° to the right of the perpendicular bisector θ0; The process of selecting 5 laser beam data from the laser rays generated at 9 angles is as follows: the scanning beam data at five angles θ0, θ1, θ2, θ3, and θ4 are used at the first position in the current concentration reconstruction area R, and the scanning beam data at five angles θ0, θ1, θ2, θ3, and θ4 are used at the second position. -1 , θ0, θ1, θ2, θ3 scanning beam data, the third position uses θ -2 ,θ -1 , θ0, θ1, θ2 scanning beam data, the fourth point θ -3 ,θ -2 ,θ -1 , θ0, θ1 scanning beam data, the fifth position θ -4 ,θ -3 ,θ -2 ,θ -1 , θ0 scanning beam data, so as to ensure that t When the concentration is within the range, the UAV has 5 valid beam data at each scanning position, and a total of 25 non-overlapping laser information that can completely cover the concentration reconstruction area R is obtained, completely covering the concentration reconstruction area R.

3. The method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology according to claim 2, characterized in that: The integrated concentrations corresponding to the 25 laser beams are transmitted to the data processing platform via wireless communication to form 25 data. The 25 data are used to form a 25-dimensional projection column vector A (ppm*m). The length of the j-th grid that the i-th beam passes through is recorded as L. i，j , by L i，j Construct a 25*16-dimensional projection matrix L(m), and define the natural gas concentration inside the j-th grid as α j , α1-α 16 The 16-dimensional reconstruction vector α is the concentration reconstruction result. The natural gas concentrations of the 16 discrete grids are represented by α1-α 16 express; Known: In short: L·α=A Where A is the projection result vector composed of the real-time updated path integral concentration information obtained during the inspection process, and L is the projection matrix known from the optical path layout. Solving the reconstruction vector α is to obtain the concentration distribution information of the natural gas in the inspection plane.

4. The method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology according to claim 3 is characterized in that: The effective path length L in each discrete grid i,j Determined by the width w of the discrete grid and the angle θ between the light and the coordinate axis; the integral concentration A corresponding to each beam path i Depend on Measured, where I0 and I t are the incident light intensity and the transmitted light intensity, P is the pressure, S(T) is the line intensity, L i,j To absorb the effective path length, the pressure and line intensity in the entire measurement area are set to constant values ​​at normal temperature and pressure.

5. The method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology according to claim 4 is characterized in that: The equation L·α=A is solved by ART algebraic iteration to calculate the gas concentration α in 16 discrete grids in the first 8m*8m concentration reconstruction area R1. j , the ART algebraic iteration process is as follows: The initial gas concentration distribution is vector α (0) Under the premise of (0) =0, the projection iteration formula for reconstructing the vector α is as follows: Where i is the number of iterations, α (i) is the concentration reconstruction result of the i-th iteration, A i is the projection result of the i-th beam, that is, the integrated concentration of natural gas on the i-th beam path, L i is the projection coefficient of the i-th light ray, that is, the i-th row vector of the matrix L, λ is the relaxation parameter, λ=0.

001. (1) Assign values ​​to the gas concentrations in the 16 discrete grids, and assign the initial value of the natural gas concentration of all discrete grids in the concentration reconstruction area R1 to 0. (i) =0, i=0,j=1,2…,16; (2) For the ith laser ray among the 25 lasers, the estimated projection result A on the ith row of the equation is calculated using the following formula: i * , that is, solving the estimated integrated concentration of natural gas on the i-th laser path: (3) The error Δ between the actual projection and the estimated projection on the i-th row of equations is calculated using the following formula: i , thus obtaining the error between the actual integrated natural gas concentration on the ith laser path and the estimated integrated natural gas concentration on the ith laser path: (4) Calculate the corrected value of the gas concentration in the jth discrete grid: (5) Use the correction value to correct the gas concentration value in the jth discrete grid: (6) Using the result of the previous round as the initial value, repeat the process (2) to (5) until the convergence condition is reached; each round of complete iteration is 25 times. If the number of iterations i>25, the next round of iteration is performed. Each round of iteration uses the result of the previous round of iteration as the initial value, that is, the initial value α of the next round of iteration (0) The value is the α of the previous round (25) , until the preset convergence condition is met, and the convergence condition meets the maximum number of iterations or the error convergence criterion: ||a (i) ||-||a (i-1) ||≤ε(ε=10 -5 ) The natural gas concentration distribution of 16 discrete grids in R1 is calculated as follows: α1-α 16 , concentration reconstruction concentration region set R'{R1, R2...R i The remaining elements in} are obtained by repeating the contents of (1)-(6).

6. The method for remotely measuring two-dimensional distribution of gas concentration based on absorption spectroscopy technology according to claim 5, characterized in that: The natural gas concentration α1-α of the 16 discrete grids corresponding to the concentration reconstruction area R1 is 16 , as a reference value to constrain some free variables in the concentration reconstruction region R2: The ART algorithm is combined with telemetry, and the strategy of local constrained variables is adopted. Part of the data solved last time is used as a reference. α1, α2, α3, and α5 in R2 are selected as the four redundant free variables to constrain. Since α1, α2, α3, and α5 in R2 and α5, α6, α7, and α9 in R1 are the same discrete grid, and the values ​​of α1, α2, α3, and α5 in R2 are equal to those of α5, α6, α7, and α9 in R1, α5, α6, α7, and α9 in R1 are used to solve the remaining 12 unknown solutions in the re-formed R2. The four free variables α1, α2, α3, and α5 in the projection equation L*α=A to be solved in R2 are constrained by α5, α6, α7, and α9 solved by R1, that is, α1, α2, α3, and α5 in R2 are replaced by α5, α6, α7, and α9 in R1. The resulting equation is: Converted into the following form: Simplified to: L R2 ×α R2 =A R2 The only unknown numbers left are α4, α6, α7, α8, α9, α 10 , α 11 , α 12 , α 13 , α 14 , α 15 , α 16 , thus forming a new unknown number of 12, the projection matrix L R2 The equation to be solved is of rank 12, and other solution values ​​in R1 can be used as initial values ​​to participate in the ART algebraic iterative reconstruction process of the projection matrix in R2; R t The concentration distribution inside is given by R t-1 The solved variable constraints within are constrained and ART algebraic iterative reconstruction is performed cyclically to complete the natural gas concentration of all concentration reconstruction areas R on the entire UAV route.

7. A drone, characterized in that: The invention comprises an unmanned aerial vehicle (UAV) body, under which a remote sensing laser transceiver for detecting gas concentration information by laser is arranged through a gimbal, and the type of gas detected can be adjusted according to actual needs. A computing device is arranged in the UAV, and the computing device comprises a processor and a memory, wherein the processor is electrically connected to the memory, and the memory is used to store instructions and data, and the processor is used to execute the method for remotely sensing two-dimensional distribution of gas concentration based on decomposition and absorption spectroscopy technology as described in any one of claims 1 to 6.

Images