Natural gas leakage laser detection interference suppression method based on multi-dimensional signal filtering

Abstract

The invention discloses a natural gas leakage laser detection interference suppression method based on multi-dimensional signal filtering, and particularly relates to the technical field of natural gas leakage laser detection. Ambient noise is removed by adopting wavelet transform and sliding mean filtering, and the signal quality is improved by combining background spectrum compensation; calculating a reflected light intensity attenuation ratio and a phase offset difference, evaluating the stability of a measured value, and adaptively adjusting a laser incident angle or increasing detection points based on an evaluation result to reduce interference influence; finally, the gas concentration is calculated through the optimized signals, trend prediction is conducted in combination with time sequence analysis, the accuracy of leakage detection is ensured, and if the concentration exceeds a set threshold value, leakage alarm is triggered in real time; according to the invention, the stability and reliability of natural gas leakage detection in a narrow space are effectively improved, false alarms and missing alarms are reduced, and a high-precision and high-robustness detection scheme is provided for gas pipe network inspection.

Description

Technical Field

[0001] This invention relates to the field of laser detection technology for natural gas leaks, and more specifically to a method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering. Background Technology

[0002] The interference suppression method for laser detection of natural gas leaks based on multidimensional signal filtering is a technique that utilizes signal processing technology to improve the accuracy of laser detection of natural gas leaks. This method reduces interference from the external environment, such as temperature fluctuations, aerosol scattering, and background noise, through multidimensional signal filtering (e.g., signal processing in the time, frequency, and spatial domains), thereby improving the sensitivity and reliability of the detection system for natural gas leaks.

[0003] The existing technology has the following shortcomings:

[0004] During urban gas pipeline network inspections, tunable diode laser absorption spectroscopy (TDS) technology is used for natural gas leak detection. While it boasts high sensitivity and rapid response, issues arise when detecting natural gas in confined spaces (such as pipe corridors and underground tunnels). In such cases, multiple reflections from metal or smooth walls can cause multipath interference, resulting in the receiver detecting multiple signals with different phases. This leads to unstable or distorted measurements. This signal superposition effect can amplify or reduce gas concentration in certain areas, affecting leak detection, causing false alarms or decreased detection accuracy, and ultimately impacting the reliability of natural gas leak monitoring. Summary of the Invention

[0005] The purpose of this invention is to provide a method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering, so as to overcome the shortcomings of the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering, comprising:

[0007] In a confined space, the detection equipment is arranged using tunable diode laser absorption spectroscopy technology, the laser emission angle and receiver position are adjusted, and wavelength and power calibration is performed.

[0008] It emits laser signals of a specific wavelength and receives the signals after they have been absorbed by the gas. It uses wavelet transform and moving mean filtering to remove environmental noise and then compensates by comparing the signal with the background spectrum.

[0009] The intensity attenuation ratio and phase shift difference of the reflected light are collected and calculated to evaluate the stability of the measured values ​​under the multipath interference effect caused by multiple laser reflections.

[0010] Based on the evaluation results, the laser incident angle may be adjusted or more detection points may be added for areas with high interference.

[0011] The gas concentration is calculated based on the optimized signal, and trend prediction is performed by combining time series analysis. If the concentration exceeds the set threshold, a leak alarm is triggered.

[0012] Preferably, wavelet transform and moving average filtering are used to remove high-frequency noise and short-term fluctuations, and background spectral compensation is combined, including: correcting errors caused by environmental interference by recording the background spectrum under leak-free conditions.

[0013] Preferably, the method for calculating the reflected light intensity attenuation ratio is as follows: Under ideal conditions of no obstructions and no multipath reflection, record the direct transmitted light intensity of the laser through the detection area; use a high-sensitivity photodetector to measure the transmitted light signal and calculate the light intensity of the direct transmission path, expressed as: I direct =P0e -αL Among them: I direct Let P0 be the light intensity of the direct transmission path, α be the absorption coefficient of the medium for the laser, and L be the length of the direct propagation path; record the received light intensity after multipath propagation, expressed as: Among them: I multi-path I represents the actual received light intensity. i R is the light intensity after the i-th reflection. i Let n be the reflectivity of the reflective surface to the laser, and n be the number of reflections the laser undergoes. Calculate the reflected light intensity attenuation ratio, which is defined as the ratio of the direct transmitted light intensity to the multipath propagating light intensity.

[0014] Preferably, the phase offset difference is calculated as follows: In natural gas leak detection, after the laser passes through the gas, the signal measured by the receiver is expressed as: x(t)=A(t)cos(ωt+φ(t))+n(t); where: x(t) is the observed light signal, A(t) is the signal amplitude, ω is the angular frequency of the laser signal, φ(t) is the instantaneous phase, and n(t) is the observation noise;

[0015] According to Bayes' theorem, given the observed data x, the posterior probability distribution of the phase φ is: Where: P(φ|x) is the posterior probability distribution of phase offset given measurement data x, P(x|φ) is the likelihood function of the observed data, P(φ) is the prior distribution, and P(x) is the normalization factor;

[0016] Assuming the noise n(t) is zero-mean Gaussian noise, the probability distribution of the observed signal x(t) is: in: To observe the variance of the noise, we assume that the phase φ follows a Gaussian prior distribution with mean φ0, i.e.: in: To calculate the prior variance, the phase offset standard deviation (PSSD) is calculated using the following formula:

[0017] Preferably, the reflected light intensity attenuation ratio and phase shift difference are normalized so that they are both between [0,1]. The stability analysis value of the measured value under the multipath interference effect is calculated based on the normalized reflected light intensity attenuation ratio and phase shift difference.

[0018] Preferably, the stability analysis value of the measured value under the multipath interference effect is compared with a predetermined threshold. If the stability analysis value of the measured value under the multipath interference effect is greater than or equal to the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is high, and the corresponding area is divided into a low interference area. If the stability analysis value of the measured value under the multipath interference effect is less than the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is low, and the corresponding area is divided into a high interference area.

[0019] Preferably, based on the evaluation results, for areas with high interference, the laser incident angle is adjusted or the number of detection points is increased, specifically including:

[0020] If the detection area is determined to be a high-interference area, the laser incident angle θ needs to be adjusted to reduce multipath interference. The formula for calculating the adjustment angle is: θ opt =θ mit +k1(1-ZF); where: θ opt For the optimized laser incident angle, θ mit is the original laser incident angle, ZF is the stability analysis value of the measured value under the multipath interference effect, and k1 is the angle adjustment coefficient, which controls the angle adjustment range;

[0021] If adjusting the incident angle still cannot effectively improve stability, then additional detection points N need to be added in the high-interference area. new To improve measurement accuracy, the formula for calculating the number of new detection points is: N new =k2(1-ZF); where: k2 is the detection point adjustment coefficient, which controls the number of newly added detection points.

[0022] Preferably, the gas concentration is calculated based on the optimized signal, and time series analysis is used to predict the trend and determine the gas leakage situation. If the early warning conditions are met, a three-level alarm mechanism is triggered, and the alarm is sent to the monitoring center. Combined with the remote control system, automatic emergency measures are executed.

[0023] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0024] This invention optimizes the layout of TDLAS equipment, adjusts the laser emission angle and receiver position, and performs wavelength and power calibration to reduce interference caused by multiple laser reflections. Wavelet transform and moving average filtering are used to remove environmental noise, combined with background spectral compensation to improve signal stability. The stability of the measured values ​​is evaluated by calculating the reflected light intensity attenuation ratio and phase shift difference. Based on the normalized stability analysis values, the laser incident angle is dynamically adjusted or detection points are added to reduce measurement errors in high-interference areas. The optimized signal is used to calculate gas concentration, and trend prediction is performed using time series analysis to ensure the accuracy of leak detection. A three-level alarm mechanism is triggered when the gas concentration exceeds a set threshold, and automated emergency response is achieved through wireless communication and a remote control system. This invention effectively improves the accuracy, stability, and intelligence level of natural gas leak detection, reduces false alarms and missed alarms, and provides a safer, more efficient, and accurate detection solution for gas pipeline network inspection. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0026] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] For examples, please refer to Figure 1 As shown in this embodiment, the method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering includes:

[0029] In a confined space, the detection equipment is arranged using tunable diode laser absorption spectroscopy technology, the laser emission angle and receiver position are adjusted, and wavelength and power calibration is performed.

[0030] It emits laser signals of a specific wavelength and receives the signals after they have been absorbed by the gas. It uses wavelet transform and moving mean filtering to remove environmental noise and then compensates by comparing the signal with the background spectrum.

[0031] The intensity attenuation ratio and phase shift difference of the reflected light are collected and calculated to evaluate the stability of the measured values ​​under the multipath interference effect caused by multiple laser reflections.

[0032] Based on the evaluation results, the laser incident angle may be adjusted or more detection points may be added for areas with high interference.

[0033] The gas concentration is calculated based on the optimized signal, and trend prediction is performed by combining time series analysis. If the concentration exceeds the set threshold, a leak alarm is triggered.

[0034] In this application, the accuracy and reliability of natural gas leak detection are improved by optimizing the layout of TDLAS equipment, adjusting the laser emission angle and receiver position, and performing wavelength and power calibration in narrow environments such as pipe galleries and underground tunnels.

[0035] Measure the dimensions of the detection area (width, height, length) to determine the optimal location for equipment installation. Analyze the wall material (e.g., metal, concrete) to assess potential laser reflection and absorption, avoiding multipath interference caused by highly reflective surfaces. Detect wind speed and airflow distribution, using ultrasonic anemometers or thermal airflow sensors to measure airflow within pipes or tunnels to determine the possible diffusion direction of gas leaks. Combine historical data or pressure monitoring systems to identify potential leak locations. Select key monitoring areas, such as pipe joints, valves, and flange connections. Choose the appropriate laser measurement mode based on environmental characteristics: Single-end measurement: The laser emitter and receiver are installed opposite each other, forming a penetrating detection method, suitable for wider passages (e.g., ≥2m). Reflective measurement: The laser emitter and receiver are on the same side, with a highly reflective mirror (e.g., gold or platinum coated mirror) installed opposite, suitable for narrow passages (e.g., <2m).

[0036] If the laser is emitted perpendicular to a smooth wall or metal surface, it may cause strong reflected signals, affecting the measurement at the receiving end. Therefore, the laser incident angle should be adjusted to deviate from the wall by at least 10° to 15° to reduce the impact of direct reflection. Arrange the laser along the airflow direction so that the leaking gas can affect the optical path to the greatest extent, improving detection sensitivity. Avoid propagating the laser along the axis of the pipe gallery or tunnel to reduce reflection interference and signal overlap.

[0037] Adjust the height of the receiver to place it in an area where leaked gas is likely to accumulate, such as 30-50cm above the ground (suitable for methane, as its density is less than air). Add a light shield or filter to reduce interference from external light sources (such as LEDs or incandescent lamps) on the laser receiver.

[0038] Because the wavelength of semiconductor lasers drifts with temperature (typical drift value: ~0.1 nm / ℃), a thermoelectric cooling (TEC) unit is installed to control the laser operating temperature within ±0.1°C. Measurements are performed at different temperatures (e.g., 0°C, 25°C, 50°C), and temperature-wavelength compensation curves are established to automatically correct for wavelength drift during actual measurements. Since the absorption characteristics of gas molecules change under different pressures, a pressure sensor is used to measure real-time pressure and correct the absorption spectrum.

[0039] If the laser power is too high, the receiver may be unable to accurately measure changes in light intensity due to signal oversaturation. Use an adjustable laser driver to adjust the laser power to achieve the optimal dynamic range of the detection equipment (typically 0.1–10 mW). If the laser power is too low, it may result in insufficient received signal strength, affecting detection sensitivity.

[0040] The system performs self-tests and final tests, including checking the signal-to-noise ratio (SNR) to ensure SNR > 30dB, thus guaranteeing signal quality. Laser alignment is checked, and the positions of the transmitter and receiver are adjusted to ensure maximum light intensity passes through the gas sample area. A small-scale release of methane gas at a known concentration (e.g., 100ppm) is conducted in the detection area, and the system response is monitored. The measured values ​​are compared with standard values ​​to ensure the measurement error is less than 2%.

[0041] By emitting lasers of a specific wavelength, receiving the signals absorbed by the gas, and using wavelet transform and moving average filtering to remove environmental noise, while also performing background spectral compensation, the accuracy and reliability of natural gas leak detection are improved.

[0042] Based on the spectral absorption characteristics of methane, a wavelength within the range of 1.645–1.665 μm was selected to ensure that the laser beam matched the absorption peak of the target gas. A tunable diode laser (TDL) was employed, allowing for dynamic wavelength adjustment during detection to perform real-time spectral scanning and improve measurement accuracy.

[0043] The laser is emitted to the detection area via a fiber optic coupler or free-space optical system. Pulse mode is suitable for remote detection and reduces background noise interference. Continuous wave mode (CW mode) is suitable for short-distance, high-precision measurements and improves the signal-to-noise ratio (SNR). Laser power is controlled (typically 0.1–10 mW) to avoid signal oversaturation or attenuation. A high-sensitivity photodetector (such as an InGaAs photodiode) is used to receive the light signal after gas absorption. Lock-in amplification technology is used to improve the detection capability of weak signals and suppress environmental noise interference. Transmittance (T) is calculated: Where: I0 is the reference light intensity (background spectrum) without gas, and I is the light intensity after gas absorption.

[0044] In confined spaces, laser signals can be affected by noise from aerosol scattering, temperature changes, and electromagnetic interference, thus requiring noise reduction. A window function is used to smooth the signal and reduce high-frequency noise. Where: y[n] is the filtered signal, x[n] is the original signal, and N is the sliding window size (usually 3 to 5).

[0045] The discrete wavelet transform (DWT) is used to decompose the signal into multi-scale components, namely low-frequency components (trend signal) and high-frequency components (noise signal). Then, high-frequency noise is removed and the signal is reconstructed. The specific process includes: selecting a suitable wavelet basis (e.g., Daubechies db4), performing wavelet decomposition to obtain signal components at different scales, setting a noise threshold (e.g., the 3σ rule), removing high-frequency noise, and performing wavelet reconstruction to obtain the denoised signal.

[0046] Under leak-free conditions, the background spectrum Ibg(λ) is recorded for subsequent compensation. Since water vapor, carbon dioxide, etc., may interfere with the spectrum, the background correction factor is calculated: Icorr(λ) = Imeas(λ) - Ibg(λ); where Imeas(λ) is the measured transmission spectrum and Icorr(λ) is the compensated spectrum.

[0047] The intensity attenuation ratio and phase shift difference of the reflected light are collected and calculated to evaluate the stability of the measurements under the multipath interference effect caused by multiple laser reflections. Specifically, this includes:

[0048] The method for calculating the reflected light intensity attenuation ratio is as follows: Under ideal conditions of no obstructions and no multipath reflections, record the intensity of the direct transmitted light through the detection area. A high-sensitivity photodetector (such as an InGaAs sensor) is used to measure the transmitted light signal, and environmental background noise is removed. The intensity of the light along the direct transmission path is calculated using the expression: I direct =P0e -αL Among them: I direct Let P0 be the light intensity along the direct transmission path, P0 be the initial laser power, α be the absorption coefficient of the medium for the laser, and L be the length of the direct propagation path. In actual measurement environments, the laser may undergo multiple reflections through walls, equipment, or pipes before reaching the receiver. The received light intensity after multipath propagation is recorded, expressed as: Among them: I multi-path I represents the actual received light intensity (including direct transmitted light + multipath reflected light). i R is the light intensity after the i-th reflection. i The reflectivity of the reflecting surface to laser light (0) <R i<1), where n is the number of reflections the laser experiences. The reflected light intensity attenuation ratio (RIAR) is calculated. The reflected light intensity attenuation ratio is defined as the ratio of the directly transmitted light intensity to the multipath propagating light intensity, expressed as:

[0049] The phase offset difference is calculated as follows: In natural gas leak detection, after the laser passes through the gas, the signal measured by the receiver is expressed as: x(t)=A(t)cos(ωt+φ(t))+n(t); where: x(t) is the observed light signal, A(t) is the signal amplitude (which may vary due to environmental influences), ω is the angular frequency of the laser signal, φ(t) is the instantaneous phase, and n(t) is the observation noise.

[0050] According to Bayes' theorem, given the observed data x, the posterior probability distribution of the phase φ is: Where: P(φ|x) is the posterior probability distribution of the phase offset given the measurement data x, P(x|φ) is the likelihood function of the observed data, P(φ) is the prior distribution, and P(x) is the normalization factor (normalization makes the sum of the probability distributions equal to 1, which does not affect the final phase estimation).

[0051] Assuming the noise n(t) is zero-mean Gaussian noise, the probability distribution of the observed signal x(t) is: in: The variance of the observed noise represents the degree of fluctuation in the measured signal. The phase φ is assumed to follow a Gaussian prior distribution with mean φ0, i.e.: in: Let be the prior variance, representing the degree of fluctuation in phase changes during historical measurements. The phase offset standard deviation (PSSD) is calculated; the phase offset difference measures the uncertainty in phase estimation, and its calculation formula is:

[0052] The reflected light intensity attenuation ratio and phase shift difference are normalized to be within the range of [0,1]. The stability analysis value of the measured value under the multipath interference effect is calculated based on the normalized reflected light intensity attenuation ratio and phase shift difference.

[0053] For example, the present invention can use the following formula to calculate the stability analysis value of the measured value under the multipath interference effect, the calculation expression is: In the formula, ZF is the stability analysis value of the measured value under the multipath interference effect, RIAR is the reflected light intensity attenuation ratio, PSSD is the phase shift standard deviation, and a1 and a2 are the weighting coefficients of the reflected light intensity attenuation ratio and the phase shift difference (which can be optimized according to experimental experience or machine learning), and a1 and a2 are both greater than 0.

[0054] The stability analysis value of the measured value under the multipath interference effect is compared with a predetermined threshold. If the stability analysis value of the measured value under the multipath interference effect is greater than or equal to the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is high, and the corresponding area is classified as a low interference area. If the stability analysis value of the measured value under the multipath interference effect is less than the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is low, and the corresponding area is classified as a high interference area.

[0055] It should be noted that the predetermined threshold can be set through experimental determination or machine learning optimization, and a reasonable predetermined threshold (such as 0.5) can be set. The predetermined threshold can also be dynamically adjusted to automatically optimize the classification criteria according to different environments.

[0056] Based on the evaluation results, the laser incident angle can be adjusted or more detection points can be added for areas with high interference.

[0057] If the detection area is determined to be a high-interference area, the laser incident angle θ needs to be adjusted to reduce multipath interference. The formula for calculating the adjustment angle is: θ opt =θ mit +k1(1-ZF); where: θ opt θ represents the optimized laser incident angle (in degrees). mit The original laser incident angle is usually set to 0°≤θ≤90°; ZF is the stability analysis value of the measured value under the multipath interference effect; k1 is the angle adjustment coefficient, which controls the angle adjustment range (empirical value, such as k1=10°).

[0058] When ZF is low (severe multipath interference), 1-ZF is close to 1, and the adjustment angle range is large. When ZF is close to the threshold (less interference), the adjustment range is small.

[0059] If adjusting the incident angle still cannot effectively improve stability, then additional detection points N need to be added in the high-interference area. new This improves measurement accuracy. The formula for calculating the number of new detection points is: N new =k2(1-ZF); where: k2 is the detection point adjustment coefficient, which controls the number of new detection points (empirical value, such as k2=5).

[0060] When ZF is low (severe multipath interference), 1-ZF approaches 1, requiring more detection points. When ZF is close to the threshold (less interference), the number of new detection points is small or remains unchanged.

[0061] The gas concentration is calculated based on the optimized signal, and trend prediction is performed by combining time series analysis. If the concentration exceeds the set threshold, a leak alarm is triggered.

[0062] After adjusting the laser incident angle or adding detection points, optimized laser signal data is obtained. To ensure the accuracy of natural gas leak detection, the optimized signal needs to be used to calculate gas concentration and combined with time series analysis for trend prediction to determine whether a leak alarm should be triggered.

[0063] The specific process is as follows:

[0064] The signal after multipath interference suppression contains transmitted light intensity information. This information is converted into an electrical signal by the photodetector at the receiving end and then filtered to remove residual noise. The acquired signal data is more stable at this point, and the influence of multipath interference on the measurement values ​​is reduced.

[0065] The absorptivity of a target gas (such as methane) is calculated using the principle of Tunable Diode Laser Absorption Spectroscopy (TDLAS). The Beer-Lambert law is then applied to convert the signal absorption intensity into a gas concentration value. Since gas concentration is closely related to environmental parameters such as temperature, humidity, and air pressure, error compensation using environmental sensor data is necessary to ensure the accuracy of the measurements.

[0066] To adapt to different environments, the system can dynamically adjust the gas concentration alarm threshold based on historical data and environmental parameters to reduce false alarms. For example, in well-ventilated environments, a higher alarm threshold can be set to avoid false alarms. In enclosed spaces (such as underground utility tunnels), where gas tends to accumulate, the alarm threshold can be appropriately lowered to improve safety.

[0067] Time series analysis for trend prediction specifically includes:

[0068] Collect and store historical concentration data: Record gas concentration values ​​at continuous intervals and establish a concentration change database. Use sliding window technology to extract the gas concentration data sequence for the most recent period.

[0069] Time series analysis is used to predict future trends: Moving average (MA) is used to smooth gas concentration data to reduce the impact of instantaneous fluctuations. Autoregressive moving average (ARIMA) or Long Short-Term Memory (LSTM) models are employed to analyze gas concentration trends and predict future concentration values. Historical data and meteorological factors (wind speed, temperature, humidity, etc.) are combined to predict the diffusion path and concentration change rate of the leaked gas.

[0070] Anomaly Detection: The current concentration change rate is calculated and compared with a normal gas diffusion model to determine if there is an abnormal leakage trend. If the concentration value increases exponentially, or the concentration change rate is significantly higher than the normal fluctuations in the background environment within a short period of time, a gas leak is suspected.

[0071] For example, the steps for predicting gas concentration trends using an LSTM model include:

[0072] Data collection: Collect and organize historical monitoring data of gas concentration, which is usually a series of continuous concentration values ​​sampled at time intervals (such as every minute or every hour).

[0073] Data cleaning: Remove outliers and missing values, and perform interpolation or smoothing to ensure data quality.

[0074] Normalization: The original gas concentration values ​​are scaled to a fixed range (e.g., 0-1) to improve the stability and efficiency of neural network training.

[0075] Sliding window partitioning: The sliding window method is used to divide continuous time series data into several samples, each containing:

[0076] Input sequence (e.g., concentration values ​​from the past 10 time points)

[0077] The corresponding output (i.e., the predicted concentration value at the next time step or several time steps in the future)

[0078] Choosing a model architecture: Construct a neural network containing one or more LSTM units. Common architectures include:

[0079] Input layer → LSTM layer → Dense fully connected layer → Output layer;

[0080] Set training parameters such as learning rate, batch size, and number of training epochs.

[0081] Model training: Use historical data to train the model, optimize prediction accuracy, and minimize the loss function (such as mean squared error, MSE).

[0082] Use a separate validation or test set to evaluate model performance.

[0083] Indicators such as mean squared error (MSE) and mean absolute error (MAE) are used to measure prediction accuracy.

[0084] If the error is large, it can be improved by adjusting the model structure, increasing the amount of data, or optimizing the hyperparameters.

[0085] Using a pre-trained LSTM model and the latest gas concentration data sequence as input, it predicts concentration values ​​at one or more future time points. It can perform rolling or multi-step predictions for real-time trend tracking.

[0086] The predicted results are compared with the set safety thresholds. If the predicted value continues to rise or exceeds the threshold, it indicates a potential risk of gas leakage, and an early warning signal is issued. Further analysis, considering the rate of concentration change and meteorological conditions, can be used to determine whether it is an abnormal pollution event.

[0087] The system triggers a leak alarm when any of the following conditions are met:

[0088] The instantaneous concentration exceeds a set threshold (e.g., 50 ppm). The rate of concentration change exceeds a set threshold (e.g., 10 ppm / min). Time series analysis predicts that the future concentration will exceed the safety limit. Multiple detection points simultaneously detect abnormal concentrations, enhancing detection reliability.

[0089] Based on gas concentration values ​​and trend prediction results, different alarm levels can be set:

[0090] Level 1 Warning (Minor Leakage): Concentration is close to the threshold, fluctuates briefly but has no obvious upward trend, and is only recorded.

[0091] Level 2 Warning (Moderate Leakage): The concentration exceeds the threshold and the rate of concentration change is relatively fast. It is recommended that inspection personnel go to check.

[0092] Level 3 Warning (Severe Leakage): If the concentration rises sharply or reaches the explosion limit (e.g., 5% LEL), an emergency alarm will be triggered immediately, relevant departments will be notified, and emergency measures will be initiated.

[0093] Once an alarm is triggered, the system can transmit the alarm information to the monitoring center via wireless communication (such as LoRa, 5G, or Wi-Fi). The monitoring system displays real-time information such as the leak location, concentration trends, and wind speed and direction to aid decision-making. If the leak occurs in a smart pipeline network environment, it can be combined with a GIS (Geographic Information System) to provide a visualized map of the leak area, facilitating rapid location and response by personnel.

[0094] Alarm follow-up handling and decision-making include: shutting off relevant valves: If the leak occurs near a gas pipeline, valves can be remotely shut off via the SCADA system to reduce gas diffusion. Activating ventilation equipment: If the leak occurs in a confined space (such as an underground tunnel), the exhaust system can be automatically activated to reduce gas concentration. Drone / robot inspection: In high-risk areas, drones or robots can be deployed for secondary inspections to confirm the leak.

[0095] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0096] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0097] It should be understood that the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be singular or plural. Additionally, the character " / " in this document generally indicates an "or" relationship between the preceding and following related objects, but it may also indicate an "and / or" relationship; please refer to the context for specific understanding. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0098] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering, characterized in that: include: In a confined space, the detection equipment is arranged using tunable diode laser absorption spectroscopy technology, the laser emission angle and receiver position are adjusted, and wavelength and power calibration is performed. It emits laser signals of a specific wavelength and receives the signals after they have been absorbed by the gas. It uses wavelet transform and moving mean filtering to remove environmental noise and then compensates by comparing the signal with the background spectrum. The intensity attenuation ratio and phase shift difference of the reflected light are collected and calculated to evaluate the stability of the measured values ​​under the multipath interference effect caused by multiple laser reflections. Based on the evaluation results, the laser incident angle may be adjusted or more detection points may be added for areas with high interference. The gas concentration is calculated based on the optimized signal, and trend prediction is performed by combining time series analysis. If the concentration exceeds the set threshold, a leak alarm is triggered.

2. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 1, characterized in that: Wavelet transform and moving mean filtering are used to remove high-frequency noise and short-term fluctuations, and background spectral compensation is combined, including: correcting errors caused by environmental interference by recording the background spectrum under leak-free conditions.

3. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 1, characterized in that: The method for calculating the reflected light intensity attenuation ratio is as follows: Under ideal conditions of no obstructions and no multipath reflections, record the direct transmitted light intensity of the laser through the detection area; use a high-sensitivity photodetector to measure the transmitted light signal and calculate the light intensity along the direct transmission path, expressed as: I direct =P0e -αL Among them: I direct Let P0 be the light intensity of the direct transmission path, α be the absorption coefficient of the medium for the laser, and L be the length of the direct propagation path; record the received light intensity after multipath propagation, expressed as: Among them: I multi-path I represents the actual received light intensity. i R is the light intensity after the i-th reflection. i Let n be the reflectivity of the reflective surface to the laser, and n be the number of reflections the laser undergoes. Calculate the reflected light intensity attenuation ratio, which is defined as the ratio of the direct transmitted light intensity to the multipath propagating light intensity.

4. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 3, characterized in that: The phase offset difference is calculated as follows: In natural gas leak detection, after the laser passes through the gas, the signal measured by the receiver is expressed as: x(t)=A(t)cos(ωt+φ(t))+n(t); where: x(t) is the observed light signal, A(t) is the signal amplitude, ω is the angular frequency of the laser signal, φ(t) is the instantaneous phase, and n(t) is the observation noise; According to Bayes' theorem, given the observed data x, the posterior probability distribution of the phase φ is: Where: P(φ|x) is the posterior probability distribution of phase offset given measurement data x, P(x|φ) is the likelihood function of the observed data, P(φ) is the prior distribution, and P(x) is the normalization factor; Assuming the noise n(t) is zero-mean Gaussian noise, the probability distribution of the observed signal x(t) is: in: To observe the variance of the noise, we assume that the phase φ follows a Gaussian prior distribution with mean φ0, i.e.: in: To calculate the prior variance, the phase offset standard deviation (PSSD) is calculated using the following formula:

5. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 4, characterized in that: The reflected light intensity attenuation ratio and phase shift difference are normalized to be within the range of [0,1]. The stability analysis value of the measured value under the multipath interference effect is calculated based on the normalized reflected light intensity attenuation ratio and phase shift difference.

6. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 5, characterized in that: The stability analysis value of the measured value under the multipath interference effect is compared with a predetermined threshold. If the stability analysis value of the measured value under the multipath interference effect is greater than or equal to the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is high, and the corresponding area is divided into a low interference area. If the stability analysis value of the measured value under the multipath interference effect is less than the predetermined threshold, it indicates that the stability of the measured value under the multipath interference effect is low, and the corresponding area is classified as a high interference area.

7. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 6, characterized in that: Based on the evaluation results, for areas with high interference, the laser incident angle can be adjusted or more detection points can be added, specifically including: If the detection area is determined to be a high-interference area, the laser incident angle θ needs to be adjusted to reduce multipath interference. The formula for calculating the adjustment angle is: θ opt =θ mit +k1(1-ZF); where: θ opt For the optimized laser incident angle, θ mit is the original laser incident angle, ZF is the stability analysis value of the measured value under the multipath interference effect, and k1 is the angle adjustment coefficient, which controls the angle adjustment range; If adjusting the incident angle still cannot effectively improve stability, then additional detection points N need to be added in the high-interference area. new To improve measurement accuracy, the formula for calculating the number of new detection points is: N new =k2(1-ZF); where: k2 is the detection point adjustment coefficient, which controls the number of newly added detection points.

8. The method for suppressing interference in laser detection of natural gas leaks based on multidimensional signal filtering according to claim 7, characterized in that: Based on the optimized signal, the gas concentration is calculated, and time series analysis is used to predict the trend and determine the gas leakage situation. If the early warning conditions are met, a three-level alarm mechanism is triggered, and the alarm is sent to the monitoring center. Combined with the remote control system, automatic emergency measures are executed.

Images