Gas pipeline leakage detection method based on unmanned aerial vehicle

Abstract

The invention discloses a gas pipeline leakage detection method based on an unmanned aerial vehicle, and the method comprises the following steps: S1, controlling the unmanned aerial vehicle to fly along a preset pipeline inspection path, and synchronously collecting the following data: S2, carrying out the real-time fusion processing of the data collected in S1, and generating a leakage probability score; s3, when the leakage probability score exceeds a preset threshold value, controlling the unmanned aerial vehicle to hover and marking a leakage point coordinate; and S4, generating an evaluation report containing the leakage aperture, the risk level and the environmental parameters, and encrypting and transmitting the evaluation report to the emergency terminal through the hybrid network. The unmanned aerial vehicle is adopted to carry the laser methane sensor, the multispectral thermal imager and the directional microphone array to detect whether the gas pipeline leaks or not, so that the labor intensity of workers is reduced, and meanwhile, the large-scale pipe network detection time can be shortened by adopting multi-machine cooperative task allocation, so that the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of gas pipeline detection, and in particular to a gas pipeline leakage detection method based on an unmanned aerial vehicle. Background Art

[0002] Gas pipelines play a vital role in the modern energy transmission system. However, due to long-term burial, environmental factors (such as soil corrosion, geological activities, etc.) or human factors (such as construction damage, etc.), gas pipelines may leak. Gas leakage not only causes energy waste, but more seriously, it can cause safety accidents such as explosions and fires, posing a huge threat to life and property.

[0003] Traditional gas pipeline leak detection methods mainly include manual inspections and ground detection equipment inspections. Manual inspections are inefficient and are limited by factors such as terrain and environment. For example, it is difficult to reach the location of the pipeline for inspection in some complex terrains. Ground detection equipment often needs to be deployed along the pipeline laying route, which is costly and less flexible. With the continuous development of drone technology, using drones for gas pipeline leak detection has become a very promising solution. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a gas pipeline leakage detection method based on drone.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a gas pipeline leakage detection method based on drone, comprising the following steps:

[0006] Step S1: Control the drone to fly along the preset pipeline inspection path and simultaneously collect the following data:

[0007] a) Detect gas concentration around the pipeline through laser methane sensor;

[0008] b) Obtain the temperature distribution on the pipeline surface through a multi-spectral thermal imager;

[0009] c) Capturing ambient sound wave signals through a directional microphone array;

[0010] Step S2: Perform real-time fusion processing on the data collected in step S1 to generate a leakage probability score;

[0011] Step S3: When the leakage probability score exceeds a preset threshold, the drone is controlled to hover and mark the coordinates of the leakage point;

[0012] Step S4: Generate an assessment report including leakage aperture, risk level, and environmental parameters, and transmit it to the emergency terminal via a hybrid network encryption.

[0013] As a further description of the above technical solution:

[0014] The step S2 comprises the following steps:

[0015] S21: Multispectral thermal imaging differential processing: pixel-level subtraction of 3.3-3.6μm and 4.2-4.5μm band data to suppress environmental thermal radiation interference;

[0016] S22: Voiceprint feature extraction: Analyze the sound wave signal through the Mel frequency cepstrum coefficient and match the pre-stored leakage voiceprint template;

[0017] S23: Data time and space synchronization: Using precise time protocol, the clock synchronization error of each sensor is ≤1 microsecond.

[0018] As a further description of the above technical solution:

[0019] The leakage probability score calculation in step S2 adopts the Dempster-Shafer evidence theory, and its basic probability distribution function is:

[0020]

[0021] Where: C CH4 is the measured value of methane concentration, C max =100ppm·m;

[0022] ΔT is the maximum temperature difference, ΔTmax = 10°C;

[0023] S audio is the voiceprint matching degree, S max =1.0;

[0024] α=0.5, β=0.3, γ=0.2 are weight coefficients.

[0025] As a further description of the above technical solution:

[0026] The step S3 includes the following operations:

[0027] S31: Start RTK-GPS and visual SLAM fusion positioning;

[0028] S32: Laser rangefinder measures the relative height of the leak point;

[0029] S33: Project a red laser cross mark onto the ground to assist manual review.

[0030] As a further description of the above technical solution:

[0031] The leakage aperture estimation in step S4 adopts a fluid mechanics model:

[0032]

[0033] Where: Q is the leakage flow, which is calculated by CFD inversion of the laser methane concentration gradient field;

[0034] ρ=0.717kg / m 3 ;

[0035] ΔP is the pressure difference between the inside and outside of the pipeline, which is estimated by the peak frequency of the acoustic signal spectrum:

[0036]

[0037] The calibration factor k=1.2 was determined by laboratory spraying experiments.

[0038] As a further description of the above technical solution:

[0039] The environmental parameter calibration step is performed before step S1, and a micro-weather station mounted on a drone is used to measure wind speed, temperature and humidity.

[0040] As a further description of the above technical solution:

[0041] The hybrid network adopts a hierarchical multi-hop ad hoc network:

[0042] 5G transmission: In areas with signal strength ≥ -90dBm, data is uploaded through the NR-U band with a peak rate of 1.2Gbps.

[0043] LoRa fallback: switch to LoRa when signal is lost, and data is transmitted in blocks and encrypted;

[0044] Resume transmission: Use TCP-like ACK mechanism to record transmission offset.

[0045] As a further description of the above technical solution:

[0046] In step S1, multi-UAV collaborative detection is implemented through a distributed scheduling system:

[0047] Region division: Divide the pipeline into N sub-regions based on the Voronoi diagram. The objective function is:

[0048]

[0049] Among them (A i ) is the area of ​​the region, (v i ) is the cruising speed of the drone, (E i ) is the remaining power;

[0050] Cross-validation: When a UAV detects a leak, DS evidence fusion is triggered.

[0051] The present invention has the following beneficial effects:

[0052] 1. Compared with the existing technology, this UAV-based gas pipeline leakage detection method uses a UAV carrying a laser methane sensor, a multi-spectral thermal imager and a directional microphone array to detect whether the gas pipeline has leaks, thereby reducing the labor intensity of workers. At the same time, the use of multi-machine collaborative task allocation can shorten the detection time of large-scale pipeline networks, thereby improving detection efficiency.

[0053] 2. Compared with the existing technology, this UAV-based gas pipeline leak detection method effectively suppresses interference such as sunlight reflection and surface radiation by adopting multi-spectral thermal imaging; it uses a laser methane sensor combined with dynamic threshold adjustment to detect tiny leaks with an aperture ≥ 0.8mm; and it uses a voiceprint matching algorithm to maintain an accuracy of more than 85% in a wind noise environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a flow chart of a gas pipeline leak detection method based on drone proposed in the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, 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] Reference Figure 1 The present invention provides a gas pipeline leakage detection method based on a drone, comprising the following steps:

[0057] Step S1: Control the drone to fly along the preset pipeline inspection path and simultaneously collect the following data:

[0058] a) The gas concentration around the pipeline is detected by a laser methane sensor, the sensor sensitivity is 0.05ppm·m, the sampling frequency is ≥100Hz, and the detection spectrum range is the methane absorption peak of 1653nm;

[0059] b) Obtain the temperature distribution on the pipeline surface through a multi-spectral thermal imager, covering the 3.3-3.6μm methane radiation band, the 4.2-4.5μm CO2 interference suppression band, and the 7.5-13μm environmental background band;

[0060] c) Capture environmental sound wave signals through a directional microphone array, focusing on identifying 3-5kHz high-frequency leakage turbulence noise;

[0061] When in use, multi-UAV collaborative detection is implemented through a distributed scheduling system: first, the benchmark path is generated, the pipeline GIS data is imported, the flight path spacing is calculated based on the pipeline centerline, and the detection area is divided at the same time: when dividing the detection area, the pipeline is divided into N sub-areas based on the Voronoi diagram, and the objective function is used: After the division is completed, the detection area is inspected using laser methane sensors, multi-spectral thermal imagers and directional microphone arrays, so that the required data can be collected and the collected data can be cross-verified. The leak can only be confirmed if the evidence from at least 3 drones is consistent.

[0062] Step S2: Perform real-time fusion processing on the data collected in step S1 to generate a leakage probability score;

[0063] S21: Multispectral thermal imaging differential processing: pixel-level subtraction of 3.3-3.6μm and 4.2-4.5μm band data to suppress environmental thermal radiation interference;

[0064] S22: Voiceprint feature extraction: Analyze the sound wave signal through the Mel frequency cepstrum coefficient and match the pre-stored leakage voiceprint template;

[0065] S23: Data time and space synchronization: Using precise time protocol, the clock synchronization error of each sensor is ≤1 microsecond.

[0066] When conducting data analysis, the Dempster-Shafer evidence theory is used, and its basic probability distribution function is:

[0067]

[0068] Where: C CH4 is the measured value of methane concentration, C max =100ppm·m;

[0069] ΔT is the maximum temperature difference, ΔTmax = 10°C;

[0070] S audio is the voiceprint matching degree, S max =1.0;

[0071] α=0.5, β=0.3, γ=0.2 are weight coefficients.

[0072] When the evidence of each sensor conflicts (K>0.5), the following correction strategy is initiated to correct the data to make it more accurate.

[0073] Step S3: When the leakage probability score exceeds the preset threshold, the drone is controlled to hover and mark the coordinates of the leakage point, and the following operations are performed:

[0074] S31: Start RTK-GPS and visual SLAM fusion positioning;

[0075] S32: Laser rangefinder measures the relative height of the leak point;

[0076] S33: Project a red laser cross mark onto the ground to assist manual review.

[0077] When in use, RTK-GPS uses the U-Blox ZED-F9P module and the visual SLAM uses the Intel RealSense D455 depth camera, and then the Kalman filter is used for fusion positioning, thereby improving the accuracy of positioning. Then, a laser rangefinder is used to scan the pipeline surface at an inclination angle of 10°, and the three-dimensional coordinates of the leakage point are inverted through triangulation. After the three-dimensional coordinates are determined, a red laser cross mark is projected onto the ground to facilitate manual review.

[0078] Step S4: Generate an assessment report including leakage aperture, risk level, and environmental parameters, and transmit it to the emergency terminal via a hybrid network encryption.

[0079] When estimating the leakage aperture, the fluid mechanics model is used:

[0080]

[0081] Where: Q is the leakage flow, which is calculated by CFD inversion of the laser methane concentration gradient field;

[0082] ρ=0.717kg / m 3 ;

[0083] ΔP is the pressure difference between the inside and outside of the pipeline, which is estimated by the peak frequency of the acoustic signal spectrum:

[0084]

[0085] The calibration factor k=1.2 was determined by laboratory spraying experiments.

[0086] The above calculation results are divided into risk levels:

[0087] (P≥1.0): Aperture d≥10mm or concentration>50ppm·m, triggering automatic alarm to the emergency platform;

[0088] (0.7≤P<1.0): Aperture 2mm≤d<10mm, record coordinates and prompt manual review; (P<0.7): Generate logs without triggering alarms.

[0089] When calibrating environmental parameters, it is necessary to execute before step S1, use a drone equipped with a micro-weather station to measure wind speed, temperature and humidity, and dynamically adjust the thresholds in real time.

[0090] The hybrid network uses a hierarchical multi-hop ad hoc network:

[0091] 5G transmission: In areas with signal strength ≥ -90dBm, data is uploaded through the NR-U band with a peak rate of 1.2Gbps.

[0092] LoRa fallback: switch to LoRa when signal is lost, and data is transmitted in blocks and encrypted;

[0093] Resume transmission: Use TCP-like ACK mechanism to record transmission offset.

[0094] In step S1, multi-UAV collaborative detection is implemented through a distributed scheduling system:

[0095] Region division: Divide the pipeline into N sub-regions based on the Voronoi diagram. The objective function is:

[0096]

[0097] Among them (A i ) is the area of ​​the region, (v i ) is the cruising speed of the drone, (E i ) is the remaining power;

[0098] Cross-validation: When a UAV detects a leak, DS evidence fusion is triggered.

[0099] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A gas pipeline leak detection method based on drone, characterized by: The following steps are involved: Step S1: Control the drone to fly along the preset pipeline inspection path and simultaneously collect the following data: a) Detect gas concentration around the pipeline through laser methane sensor; b) Obtain the temperature distribution on the pipeline surface through a multi-spectral thermal imager; c) Capturing ambient sound wave signals through a directional microphone array; Step S2: Perform real-time fusion processing on the data collected in step S1 to generate a leakage probability score; Step S3: When the leakage probability score exceeds a preset threshold, the drone is controlled to hover and mark the coordinates of the leakage point; Step S4: Generate an assessment report including leakage aperture, risk level, and environmental parameters, and transmit it to the emergency terminal via a hybrid network encryption.

2. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The step S2 comprises the following steps: S21: Multispectral thermal imaging differential processing: pixel-level subtraction of 3.3-3.6μm and 4.2-4.5μm band data to suppress environmental thermal radiation interference; S22: Voiceprint feature extraction: Analyze the sound wave signal through the Mel frequency cepstrum coefficient and match the pre-stored leakage voiceprint template; S23: Data time and space synchronization: Using precise time protocol, the clock synchronization error of each sensor is ≤1 microsecond.

3. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The leakage probability score calculation in step S2 adopts the Dempster-Shafer evidence theory, and its basic probability distribution function is: Where: CCH4 is the measured value of methane concentration, Cmax = 100 ppm·m; ΔT is the maximum temperature difference, ΔTmax = 10°C; Saudio is the voiceprint matching degree, Smax = 1.0; α=0.5, β=0.3, γ=0.2 are weight coefficients.

4. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The step S3 includes the following operations: S31: Start RTK-GPS and visual SLAM fusion positioning; S32: Laser rangefinder measures the relative height of the leak point; S33: Project a red laser cross mark onto the ground to assist manual review.

5. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The leakage aperture estimation in step S4 adopts a fluid mechanics model: Where: Q is the leakage flow, which is calculated by CFD inversion of the laser methane concentration gradient field; p=0.717kg / m 3 ; ΔP is the pressure difference between the inside and outside of the pipeline, which is estimated by the peak frequency of the acoustic signal spectrum: The calibration factor k=1.2 was determined by laboratory spraying experiments.

6. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The environmental parameter calibration step is performed before step S1, and a micro-weather station mounted on a drone is used to measure wind speed, temperature and humidity.

7. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: The hybrid network adopts a hierarchical multi-hop ad hoc network: 5G transmission: In areas with signal strength ≥ -90dBm, data is uploaded through the NR-U band with a peak rate of 1.2Gbps. LoRa fallback: switch to LoRa when signal is lost, and data is transmitted in blocks and encrypted; Resume transmission: Use TCP-like ACK mechanism to record transmission offset.

8. The method for detecting gas pipeline leakage based on drone according to claim 1, characterized in that: In step S1, multi-UAV collaborative detection is implemented through a distributed scheduling system: Region division: Divide the pipeline into N sub-regions based on the Voronoi diagram. The objective function is: Where (Ai) is the area of ​​the region, (vi) is the cruising speed of the drone, and (Ei) is the remaining power; Cross-validation: When a UAV detects a leak, DS evidence fusion is triggered.

Images