Gas leakage identification method and system, terminal equipment and storage medium

Abstract

The invention discloses a gas leakage identification method and system, terminal equipment and a storage medium, and the method comprises the steps: carrying out the feature extraction processing of sound data through a frequency domain feature extraction algorithm and a fuzzy entropy calculation algorithm, and obtaining the frequency domain feature and the fuzzy entropy feature of the sound data; leakage probability value prediction is carried out on the characteristic data through a gas leakage probability prediction model, and then a gas leakage result is determined according to a leakage probability prediction value, namely, the characteristic that frequency characteristics can reflect the characteristics of gas leakage in a frequency domain and the characteristic that fuzzy entropy characteristics can capture nonlinear and complexity characteristics in signals are utilized; reliable data support is provided for a subsequent prediction model, the gas leakage probability prediction model trained by sound data under different environmental conditions is utilized, the resistance of the model to the influence of environmental factors during probability prediction is improved, and the accuracy of gas leakage identification is improved. The problem that a gas leakage result cannot be accurately obtained due to interference of environmental factors in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of gas leakage identification, and in particular to a gas leakage identification method, system, terminal equipment and storage medium. Background Art

[0002] Gas leakage detection is of great significance in industrial production, storage and transportation. Gas leakage not only causes waste of resources, but also may cause major safety accidents such as fire and explosion, resulting in casualties and environmental pollution. Therefore, detection technology that can accurately detect whether gas leakage occurs is needed.

[0003] Gas leak detection technology collects changes in physical quantities in the detection environment and uses environmental changes accompanying gas leaks to identify leaks. Currently, ultrasonic detection technology is usually used to detect ultrasonic signals based on the collected sound data to determine whether a gas leak has occurred. This technology is capable of detecting tiny gas leaks. However, although this technology can detect and identify gas leaks based on the ultrasonic signals generated when a gas leak occurs, ultrasonic signals are easily interfered with by environmental factors and cannot accurately detect and identify whether a gas leak has occurred. For example, in a complex industrial environment, tiny vibrations of surrounding equipment may also generate ultrasonic signals of similar frequencies. Summary of the invention

[0004] The present invention provides a gas leakage identification method, system, terminal equipment and storage medium, which can solve the problem that the existing gas leakage detection technology is easily disturbed by environmental factors and cannot accurately detect and identify whether a gas leakage occurs.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a gas leakage identification method, comprising:

[0006] Obtain real-time sound sequence data of the area to be tested;

[0007] The frequency domain features of the real-time sound sequence data of the area to be tested are extracted by using a frequency domain feature extraction algorithm to obtain the frequency domain features of the real-time sound sequence data of the area to be tested;

[0008] The fuzzy entropy calculation algorithm is used to calculate the real-time sound sequence data of the area to be measured, and the fuzzy entropy characteristics of the real-time sound sequence data of the area to be measured are obtained;

[0009] The frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested are input into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training;

[0010] The gas leakage result of the area to be tested is determined according to the predicted value of the gas leakage probability of the area to be tested.

[0011] Furthermore, after acquiring the real-time sound sequence data of the area to be tested, the method further includes:

[0012] The real-time sound sequence data of the area to be tested is subjected to audio preprocessing to obtain processed real-time sound sequence data of the area to be tested; wherein the audio preprocessing includes denoising processing, filtering processing and normalization processing.

[0013] Furthermore, the method of extracting frequency domain features from the real-time sound sequence data of the area to be measured by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be measured includes:

[0014] The processed real-time sound sequence data of the area to be tested is subjected to frequency domain conversion and peak frequency feature extraction through Fourier transform algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested.

[0015] Furthermore, the fuzzy entropy calculation is performed on the real-time sound sequence data of the area to be measured by using a fuzzy entropy calculation algorithm to obtain the fuzzy entropy features of the real-time sound sequence data of the area to be measured, including:

[0016] The real-time sound sequence data of the area to be tested is vector-sequence constructed by a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the sound vector sequence satisfies the preset same similarity tolerance and different phase space dimensions;

[0017] The fuzzy distance between two vectors in the sound vector sequence is calculated by using a fuzzy distance calculation formula to obtain a number of vector fuzzy distances;

[0018] According to a plurality of vector fuzzy distances, the fuzzy membership of each vector fuzzy distance is calculated by a fuzzy membership function to obtain a plurality of vector fuzzy memberships;

[0019] According to the fuzzy memberships of several vectors, the average fuzzy membership of each vector is calculated by using the average fuzzy membership function to obtain the average fuzzy memberships of several vectors;

[0020] According to several vector average fuzzy memberships, the logarithmic average fuzzy membership of all vector average fuzzy memberships is calculated by the logarithmic average fuzzy membership calculation formula, and the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance is obtained;

[0021] The fuzzy entropy calculation formula is used to calculate the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance, and the fuzzy entropy characteristics of the real-time sound sequence data of the area to be measured are obtained.

[0022] Furthermore, the phase space reconstruction formula, the fuzzy distance calculation formula, the fuzzy membership function, the average fuzzy membership function, the logarithmic average fuzzy membership calculation formula and the fuzzy entropy calculation formula are as follows:

[0023] The phase space reconstruction formula is:

[0024] X(i)={u(i),u(i+1),...,u(i+m-1)}-u 0 (i), i=1, 2,..., N-m+1;

[0025] Wherein, X(i) represents the sound vector of the i-th sound data of the sound sequence in the phase space; u(i) represents the i-th sound data of the sound sequence; i represents the sampling point of the sound data; N represents the total number of sampling points; m represents the preset phase space dimension; u 0 (i) represents the average state metric of the i-th sound data in the local area,

[0026] The fuzzy distance calculation formula is:

[0027]

[0028] Among them, the represents the vector fuzzy distance in the phase space dimension m; k represents the index variable; x(i) and x(j) represent the i-th and j-th sound vectors respectively; u 0 (i) and u 0 (j) represent the average state metrics of the i-th and j-th sound data in the local area respectively;

[0029] The fuzzy membership function is:

[0030]

[0031] Among them, the represents the fuzzy membership of the vector in the phase space dimension m; a, b and c represent the first preset parameter, the second preset parameter and the third preset parameter for calculating the fuzzy membership respectively; represents the vector fuzzy distance in the phase space dimension m;

[0032] The average fuzzy membership function is:

[0033]

[0034] Among them, the represents the average fuzzy membership of the vector under the similarity tolerance; represents the fuzzy membership of the vector in the phase space dimension m;

[0035] The logarithmic mean fuzzy membership calculation formula is:

[0036]

[0037] Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; represents the average fuzzy membership of the vector under similarity tolerance;

[0038] The fuzzy entropy calculation formula is:

[0039]

[0040] Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; It represents the logarithmic mean fuzzy membership under the similarity tolerance r and phase space dimension m+1.

[0041] Furthermore, the gas leakage probability prediction model is a gas leakage probability prediction model based on BP neural network;

[0042] The model training of the gas leakage probability prediction model includes:

[0043] Obtain historical sound sequence data with actual values ​​of gas leakage probability under different environmental conditions in the area to be tested;

[0044] The frequency domain feature extraction algorithm and fuzzy entropy calculation algorithm are used to extract the frequency domain features and calculate the fuzzy entropy of the historical sound sequence data with the actual value of the gas leakage probability, and the frequency domain features and fuzzy entropy features of the historical sound sequence data are obtained;

[0045] According to the frequency domain characteristics and fuzzy entropy characteristics of the historical sound sequence data, feature extraction and nonlinear transformation activation are performed through several neurons and activation functions in several hidden layers of the gas leakage probability prediction model based on BP neural network to obtain the activated gas leakage characteristics;

[0046] The gas leakage probability prediction value is calculated for the activated gas leakage feature through the output layer of the gas leakage probability prediction model based on the BP neural network to obtain the gas leakage probability prediction value;

[0047] The predicted value of gas leakage probability and the actual value of gas leakage probability are compared to calculate the loss value. According to the loss value, the model parameters of the gas leakage probability prediction model based on BP neural network are optimized until the loss value converges to obtain a trained gas leakage probability prediction model.

[0048] Based on the above method embodiment, the present invention provides a corresponding system embodiment;

[0049] An embodiment of the present invention provides a gas leakage identification system, including: a data acquisition module, a frequency domain feature extraction module, a fuzzy entropy feature extraction module, a gas leakage probability prediction module and a gas leakage result determination module;

[0050] The data acquisition module is used to acquire real-time sound sequence data of the area to be tested;

[0051] The frequency domain feature extraction module is used to extract frequency domain features of the real-time sound sequence data of the area to be tested by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested;

[0052] The fuzzy entropy feature extraction module is used to perform fuzzy entropy calculation on the real-time sound sequence data of the area to be measured by a fuzzy entropy calculation algorithm to obtain the fuzzy entropy features of the real-time sound sequence data of the area to be measured;

[0053] The gas leakage probability prediction module is used to input the frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein, the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training;

[0054] The gas leakage result determination module is used to determine the gas leakage result of the area to be tested according to the predicted value of the gas leakage probability of the area to be tested.

[0055] Furthermore, the fuzzy entropy feature extraction module includes: a vector sequence construction unit, a vector fuzzy distance calculation unit, a vector fuzzy membership unit, a vector average fuzzy membership unit, a logarithmic average fuzzy membership unit and a fuzzy entropy feature generation unit;

[0056] The vector sequence construction unit is used to construct a vector sequence for the real-time sound sequence data of the area to be tested by using a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the vector sequence satisfies the same preset similarity tolerance and different phase space dimensions;

[0057] The vector fuzzy distance calculation unit is used to calculate the fuzzy distance between two vectors in the sound vector sequence by using a fuzzy distance calculation formula to obtain a plurality of vector fuzzy distances;

[0058] The vector fuzzy membership unit is used to calculate the fuzzy membership of each vector fuzzy distance according to a plurality of vector fuzzy distances through a fuzzy membership function to obtain a plurality of vector fuzzy memberships;

[0059] The vector average fuzzy membership unit is used to calculate the average fuzzy membership of each vector according to the plurality of vector fuzzy memberships through the average fuzzy membership function to obtain the plurality of vector average fuzzy memberships;

[0060] The logarithmic average fuzzy membership unit is used to calculate the logarithmic average fuzzy membership of all vector average fuzzy memberships according to a plurality of vector average fuzzy memberships through a logarithmic average fuzzy membership calculation formula, and obtain the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance;

[0061] The fuzzy entropy feature generating unit is used to perform fuzzy entropy calculation on the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance through a fuzzy entropy calculation formula to obtain the fuzzy entropy feature of the real-time sound sequence data of the area to be measured.

[0062] Based on the above-mentioned method embodiment, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a gas leakage identification method as described in the present invention is implemented.

[0063] Based on the above method embodiment, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which controls the device where the computer-readable storage medium is located to execute a gas leak identification method as described in the present invention when the computer program is running.

[0064] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0065] The present invention performs feature extraction processing on the sound data of the area to be tested respectively through a frequency domain feature extraction algorithm and a fuzzy entropy calculation algorithm, respectively obtains the frequency domain features and fuzzy entropy features of the sound data of the area to be tested, and predicts the leakage probability of the frequency features and the fuzzy features through a gas leakage probability prediction model to obtain a gas leakage probability prediction value of the area to be tested, and then determines the gas leakage result of the area to be tested according to the gas leakage probability prediction value, that is, the present invention utilizes the characteristics of the frequency features that can reflect the characteristics of gas leakage in the frequency domain and the characteristics of the fuzzy entropy features that can capture the nonlinear and complex characteristics in the signal, and can accurately extract the features for judging gas leakage, and provide reliable data support for subsequent gas leakage identification according to the feature data, and utilizes the gas leakage probability prediction model trained with sample data under different environmental conditions to improve the resistance of the gas leakage probability prediction model to the influence of environmental factors when performing gas leakage probability prediction, thereby improving the accuracy of gas leakage identification detection, and solving the problem that the existing gas leakage detection technology is easily interfered by environmental factors and cannot accurately detect and identify whether a gas leakage occurs. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 : A flow chart of the steps of a gas leakage identification method provided by an embodiment of the present invention;

[0067] Figure 2 : A system structure diagram of a gas leakage identification system provided by an embodiment of the present invention;

[0068] Figure 3 : Leakage probability diagram of experimental results provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0069] 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.

[0070] In the description of the present invention, it should be understood that the terms "first", "second" and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features.

[0071] Embodiment 1:

[0072] Reference Figure 1 , is a flow chart of a method for identifying a gas leak provided by an embodiment of the present invention, the method at least comprising the following steps:

[0073] Step S1: Acquire real-time sound sequence data of the area to be tested;

[0074] In this embodiment, a plurality of sound sensors are arranged in a preset manner to form a sound sensor array, and the sound sensor array is deployed in the area to be tested, and real-time sound sequence data is collected from the area to be tested according to a preset sampling frequency and a preset number of sampling points; wherein the sound sensor includes but is not limited to an analog MEMS microphone; the preset sampling frequency and the number of sampling points can be set and adjusted by a technician according to actual conditions, and illustratively, the sampling frequency is set to 300KHz, and the number of collection points is set to 1500.

[0075] In this embodiment, after obtaining the real-time sound sequence data of the area to be tested, it also includes: performing audio preprocessing on the real-time sound sequence data of the area to be tested to obtain processed real-time sound sequence data of the area to be tested; wherein the audio preprocessing includes but is not limited to denoising processing, filtering processing and normalization processing.

[0076] Step S2: extracting frequency domain features from the real-time sound sequence data of the area to be measured by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be measured;

[0077] In this embodiment, the frequency domain feature extraction is performed on the real-time sound sequence data of the area to be measured by using a frequency domain feature extraction algorithm to obtain the frequency domain features of the real-time sound sequence data of the area to be measured, including:

[0078] The processed real-time sound sequence data of the area to be tested is subjected to frequency domain conversion and peak frequency feature extraction through Fourier transform algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested.

[0079] In this embodiment, the frequency domain feature extraction algorithm is not limited to the Fourier transform algorithm. The frequency domain feature of the real-time sound sequence data of the measured area can also be extracted by the wavelet transform algorithm to obtain the frequency domain features of the real-time sound sequence data of the measured area.

[0080] Step S3: performing fuzzy entropy calculation on the real-time sound sequence data of the area to be measured by using a fuzzy entropy calculation algorithm to obtain fuzzy entropy features of the real-time sound sequence data of the area to be measured;

[0081] In this embodiment, the fuzzy entropy calculation is performed on the real-time sound sequence data of the area to be measured by using a fuzzy entropy calculation algorithm to obtain the fuzzy entropy features of the real-time sound sequence data of the area to be measured, including:

[0082] The real-time sound sequence data of the area to be tested is vector-sequence constructed by a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the sound vector sequence satisfies the preset same similarity tolerance and different phase space dimensions;

[0083] In this embodiment, the similarity tolerance r and the phase space dimension m can be set by the technician according to the actual situation. For example, the initial value of the phase space dimension can be set to 2, and the similarity tolerance r = (0.1-0.25) * SD u , S.D. u is the empirical standard deviation obtained from the statistical analysis of historical gas leakage acoustic signals.

[0084] In this embodiment, the phase space reconstruction formula is:

[0085] X(i)={u(i),u(i+1),...,u(i+m-1)}-u 0 (i), i=1, 2,..., N-m+1;

[0086] Wherein, X(i) represents the sound vector of the i-th sound data of the sound sequence in the phase space; u(i) represents the i-th sound data of the sound sequence; i represents the sampling point of the sound data; N represents the total number of sampling points; m represents the preset phase space dimension; u 0 (i) represents the average state metric of the i-th sound data in the local area,

[0087] The fuzzy distance between two vectors in the sound vector sequence is calculated by using a fuzzy distance calculation formula to obtain a number of vector fuzzy distances;

[0088] In this embodiment, the fuzzy distance calculation formula is:

[0089]

[0090] Among them, the represents the vector fuzzy distance in the phase space dimension m; k represents the index variable; x(i) and x(j) represent the i-th and j-th sound vectors respectively; u 0 (i) and u 0 (j) represent the average state metrics of the i-th and j-th sound data in the local area respectively;

[0091] According to a plurality of vector fuzzy distances, the fuzzy membership of each vector fuzzy distance is calculated by a fuzzy membership function to obtain a plurality of vector fuzzy memberships;

[0092] In this embodiment, the fuzzy membership function is:

[0093]

[0094] Among them, the represents the fuzzy membership of the vector in the phase space dimension m; a, b and c represent the first preset parameter, the second preset parameter and the third preset parameter for calculating the fuzzy membership respectively; represents the vector fuzzy distance in the phase space dimension m;

[0095] In this embodiment, the fuzzy membership function is an S-shaped curve, and parameter a can affect the opening size of the fuzzy membership function curve, and its function is to determine the relative size of the leakage and non-leakage fuzzy entropy values; parameter b can affect the steepness of the fuzzy membership function curve. When the fuzzy entropy values ​​of the leakage and non-leakage signals are easy to distinguish, parameter b can be fine-tuned upward to increase the distinction; if the fuzzy entropy values ​​of the leakage and non-leakage signals are close, parameter b is adjusted downward to reduce the error; parameter c is mainly used to distinguish the critical value, and its function is to determine the translation of the fuzzy entropy curve; the parameter values ​​of a, b and c can be set by technicians according to actual conditions.

[0096] According to the fuzzy memberships of several vectors, the average fuzzy membership of each vector is calculated by using the average fuzzy membership function to obtain the average fuzzy memberships of several vectors;

[0097] In this embodiment, the average fuzzy membership function is:

[0098]

[0099] Among them, the represents the average fuzzy membership of the vector under the similarity tolerance; represents the fuzzy membership of the vector in the phase space dimension m;

[0100] According to several vector average fuzzy memberships, the logarithmic average fuzzy membership of all vector average fuzzy memberships is calculated by the logarithmic average fuzzy membership calculation formula, and the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance is obtained;

[0101] In this embodiment, the logarithmic mean fuzzy membership calculation formula is:

[0102]

[0103] Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; represents the average fuzzy membership of the vector under similarity tolerance;

[0104] The fuzzy entropy calculation formula is used to calculate the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance, and the fuzzy entropy characteristics of the real-time sound sequence data of the area to be measured are obtained;

[0105] In this embodiment, the fuzzy entropy calculation formula is:

[0106]

[0107] Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; It represents the logarithmic mean fuzzy membership under the similarity tolerance r and phase space dimension m+1.

[0108] Step S4: inputting the frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein, the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training;

[0109] In this embodiment, the gas leakage probability prediction model may be a gas leakage probability prediction model based on a BP neural network; the model training of the gas leakage probability prediction model includes:

[0110] Obtain historical sound sequence data with actual values ​​of gas leakage probability under different environmental conditions in the area to be tested;

[0111] The frequency domain feature extraction algorithm and fuzzy entropy calculation algorithm are used to extract the frequency domain features and calculate the fuzzy entropy of the historical sound sequence data with the actual value of the gas leakage probability, and the frequency domain features and fuzzy entropy features of the historical sound sequence data are obtained;

[0112] According to the frequency domain characteristics and fuzzy entropy characteristics of the historical sound sequence data, feature extraction and nonlinear transformation activation are performed through several neurons and activation functions in several hidden layers of the gas leakage probability prediction model based on BP neural network to obtain the activated gas leakage characteristics;

[0113] In this embodiment, the number of hidden layers and the number of neurons in each hidden layer can be set by technicians according to actual conditions;

[0114] The gas leakage probability prediction value is calculated by the output layer of the gas leakage probability prediction model based on the BP neural network for the activated gas leakage feature to obtain the gas leakage probability prediction value;

[0115] In this embodiment, the types of gas leakage probability prediction values ​​output by the output layer include gas non-leakage probability, gas uncertain leakage probability and gas leakage probability;

[0116] The predicted value of gas leakage probability and the actual value of gas leakage probability are compared to calculate the loss value, and the model parameters of the gas leakage probability prediction model based on BP neural network are optimized according to the loss value until the loss value converges to obtain a trained gas leakage probability prediction model;

[0117] In this embodiment, the loss function for calculating the loss value includes but is not limited to the mean square error function and the mean absolute percentage error function; the method for optimizing the parameters of the future carbon emission prediction model to be trained includes but is not limited to the gradient descent method and the Adam optimizer.

[0118] Step S5: Determine the gas leakage result of the area to be tested according to the predicted value of the gas leakage probability of the area to be tested.

[0119] In this embodiment, refer to Figure 3 , which is a leakage probability diagram of the experimental results provided in an embodiment of the present invention, determines the gas leakage result of the tested area according to the gas leakage type corresponding to the maximum gas leakage probability prediction value; wherein the types of gas leakage probability prediction values ​​include gas non-leakage probability, gas uncertain leakage probability and gas leakage probability; the sum of the probability values ​​of gas non-leakage probability, gas uncertain leakage probability and gas leakage probability is 1.

[0120] Exemplarily, when the probability prediction value of the type of probability of no gas leakage is 0.1, the probability prediction value of the type of probability of uncertain gas leakage is 0.3, and the probability prediction value of the type of probability of gas leakage is 0.6, the result is obtained that there is a gas leakage in the test area, and a gas leakage warning is initiated; when the probability prediction value of the type of probability of no gas leakage is 0.6, the probability prediction value of the type of probability of uncertain gas leakage is 0.2, and the probability prediction value of the type of probability of gas leakage is 0.2, the result is obtained that there is no gas leakage in the test area, and no gas leakage warning is initiated; when the probability prediction value of the type of probability of no gas leakage is 0.3, the probability prediction value of the type of probability of uncertain gas leakage is 0.6, and the probability prediction value of the type of probability of gas leakage is 0.1, the result is obtained that there is an uncertain gas leakage in the test area, and a gas leakage detection suggestion is initiated.

[0121] Embodiment 2:

[0122] Reference Figure 2, is a system structure diagram of a gas leakage identification system provided by an embodiment of the present invention, the system at least includes: a data acquisition module, a frequency domain feature extraction module, a fuzzy entropy feature extraction module, a gas leakage probability prediction module and a gas leakage result determination module;

[0123] The data acquisition module is used to acquire real-time sound sequence data of the area to be tested;

[0124] The frequency domain feature extraction module is used to extract frequency domain features of the real-time sound sequence data of the area to be tested by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested;

[0125] The fuzzy entropy feature extraction module is used to perform fuzzy entropy calculation on the real-time sound sequence data of the area to be measured by a fuzzy entropy calculation algorithm to obtain the fuzzy entropy features of the real-time sound sequence data of the area to be measured;

[0126] The gas leakage probability prediction module is used to input the frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein, the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training;

[0127] The gas leakage result determination module is used to determine the gas leakage result of the area to be tested according to the predicted value of the gas leakage probability of the area to be tested.

[0128] In this embodiment, the fuzzy entropy feature extraction module includes: a vector sequence construction unit, a vector fuzzy distance calculation unit, a vector fuzzy membership unit, a vector average fuzzy membership unit, a logarithmic average fuzzy membership unit and a fuzzy entropy feature generation unit;

[0129] The vector sequence construction unit is used to construct a vector sequence for the real-time sound sequence data of the area to be tested by using a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the vector sequence satisfies the same preset similarity tolerance and different phase space dimensions;

[0130] The vector fuzzy distance calculation unit is used to calculate the fuzzy distance between two vectors in the sound vector sequence by using a fuzzy distance calculation formula to obtain a plurality of vector fuzzy distances;

[0131] The vector fuzzy membership unit is used to calculate the fuzzy membership of each vector fuzzy distance according to a plurality of vector fuzzy distances through a fuzzy membership function to obtain a plurality of vector fuzzy memberships;

[0132] The vector average fuzzy membership unit is used to calculate the average fuzzy membership of each vector according to the plurality of vector fuzzy memberships through the average fuzzy membership function to obtain the plurality of vector average fuzzy memberships;

[0133] The logarithmic average fuzzy membership unit is used to calculate the logarithmic average fuzzy membership of all vector average fuzzy memberships according to a plurality of vector average fuzzy memberships through a logarithmic average fuzzy membership calculation formula, and obtain the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance;

[0134] The fuzzy entropy feature generating unit is used to perform fuzzy entropy calculation on the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance through a fuzzy entropy calculation formula to obtain the fuzzy entropy feature of the real-time sound sequence data of the area to be measured.

[0135] Based on the above method embodiment, another embodiment is provided;

[0136] Another embodiment of the present invention provides a gas leakage identification terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a gas leakage identification method described in any one of the above method embodiments of the present invention is implemented.

[0137] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the gas leak identification terminal device.

[0138] The gas leak identification terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The gas leak identification terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art may understand that, for example, the gas leak identification terminal device may also include an input / output device, a network access device, a bus, etc.

[0139] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the gas leak identification terminal device, and uses various interfaces and lines to connect the various parts of the entire gas leak identification terminal device.

[0140] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the gas leak identification terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0141] Based on the above method embodiment, another embodiment is provided;

[0142] A storage medium provided in another embodiment of the present invention includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute a gas leakage identification method described in any one of the above method embodiments of the present invention.

[0143] Wherein, the above-mentioned storage medium is a computer-readable storage medium. The module / unit integrated in the gas leak identification system / terminal device, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0144] It should be noted that the above-mentioned terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the above-mentioned terminal device is merely an example and does not constitute a limitation on the terminal device. It may include more or fewer components, or a combination of certain components, or different components.

[0145] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A gas leakage identification method, characterized in that: include: Obtain real-time sound sequence data of the area to be tested; The frequency domain features of the real-time sound sequence data of the area to be tested are extracted by using a frequency domain feature extraction algorithm to obtain the frequency domain features of the real-time sound sequence data of the area to be tested; The fuzzy entropy calculation algorithm is used to calculate the real-time sound sequence data of the area to be measured, and the fuzzy entropy characteristics of the real-time sound sequence data of the area to be measured are obtained; The frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested are input into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training; The gas leakage result of the area to be tested is determined according to the predicted value of the gas leakage probability of the area to be tested.

2. A gas leakage identification method according to claim 1, characterized in that: After acquiring the real-time sound sequence data of the area to be tested, the method further includes: The real-time sound sequence data of the area to be tested is subjected to audio preprocessing to obtain processed real-time sound sequence data of the area to be tested; wherein the audio preprocessing includes denoising processing, filtering processing and normalization processing.

3. A gas leakage identification method according to claim 2, characterized in that: The method of extracting frequency domain features from the real-time sound sequence data of the area to be measured by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be measured includes: The processed real-time sound sequence data of the area to be tested is subjected to frequency domain conversion and peak frequency feature extraction through Fourier transform algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested.

4. A gas leakage identification method according to claim 3, characterized in that: The method of performing fuzzy entropy calculation on the real-time sound sequence data of the area to be measured by using a fuzzy entropy calculation algorithm to obtain fuzzy entropy features of the real-time sound sequence data of the area to be measured includes: The real-time sound sequence data of the area to be tested is vector-sequence constructed by a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the sound vector sequence satisfies the preset same similarity tolerance and different phase space dimensions; The fuzzy distance between two vectors in the sound vector sequence is calculated by using a fuzzy distance calculation formula to obtain a number of vector fuzzy distances; According to a plurality of vector fuzzy distances, the fuzzy membership of each vector fuzzy distance is calculated by a fuzzy membership function to obtain a plurality of vector fuzzy memberships; According to the fuzzy memberships of several vectors, the average fuzzy membership of each vector is calculated by using the average fuzzy membership function to obtain the average fuzzy memberships of several vectors; According to several vector average fuzzy memberships, the logarithmic average fuzzy membership of all vector average fuzzy memberships is calculated by the logarithmic average fuzzy membership calculation formula, and the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance is obtained; The fuzzy entropy calculation formula is used to calculate the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance, and the fuzzy entropy characteristics of the real-time sound sequence data of the area to be measured are obtained.

5. A gas leakage identification method according to claim 4, characterized in that: The phase space reconstruction formula, the fuzzy distance calculation formula, the fuzzy membership function, the average fuzzy membership function, the logarithmic average fuzzy membership calculation formula and the fuzzy entropy calculation formula are as follows: The phase space reconstruction formula is: X(i)={u(i),u(i+1),...,u(i+m-1)}-u0(i), i=1,2,...,N-m+1; Wherein, X(i) represents the sound vector of the i-th sound data of the sound sequence in the phase space; u(i) represents the i-th sound data of the sound sequence; i represents the sampling point of the sound data; N represents the total number of sampling points; m represents the preset phase space dimension; u0(i) represents the average state metric of the i-th sound data in the local area, The fuzzy distance calculation formula is: Among them, the represents the vector fuzzy distance under the phase space dimension m; k represents the index variable; x(i) and x(j) represent the i-th and j-th sound vectors respectively; u0(i) and u0(j) represent the average state measurement of the i-th and j-th sound data in the local area respectively; The fuzzy membership function is: Among them, the represents the fuzzy membership of the vector in the phase space dimension m; a, b and c represent the first preset parameter, the second preset parameter and the third preset parameter for calculating the fuzzy membership respectively; represents the vector fuzzy distance in the phase space dimension m; The average fuzzy membership function is: Among them, the represents the average fuzzy membership of the vector under the similarity tolerance; represents the fuzzy membership of the vector in the phase space dimension m; The logarithmic mean fuzzy membership calculation formula is: Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; represents the average fuzzy membership of the vector under similarity tolerance; The fuzzy entropy calculation formula is: Among them, the represents the logarithmic average fuzzy membership under the similarity tolerance r and phase space dimension m; It represents the logarithmic mean fuzzy membership under the similarity tolerance r and phase space dimension m+1.

6. A gas leakage identification method according to claim 5, characterized in that: The gas leakage probability prediction model is a gas leakage probability prediction model based on BP neural network; The model training of the gas leakage probability prediction model includes: Obtain historical sound sequence data with actual values ​​of gas leakage probability under different environmental conditions in the area to be tested; The frequency domain feature extraction algorithm and fuzzy entropy calculation algorithm are used to extract the frequency domain features and calculate the fuzzy entropy of the historical sound sequence data with the actual value of the gas leakage probability, and the frequency domain features and fuzzy entropy features of the historical sound sequence data are obtained; According to the frequency domain characteristics and fuzzy entropy characteristics of the historical sound sequence data, feature extraction and nonlinear transformation activation are performed through several neurons and activation functions in several hidden layers of the gas leakage probability prediction model based on BP neural network to obtain the activated gas leakage characteristics; The gas leakage probability prediction value is calculated for the activated gas leakage feature through the output layer of the gas leakage probability prediction model based on the BP neural network to obtain the gas leakage probability prediction value; The predicted value of gas leakage probability and the actual value of gas leakage probability are compared to calculate the loss value. According to the loss value, the model parameters of the gas leakage probability prediction model based on BP neural network are optimized until the loss value converges to obtain a trained gas leakage probability prediction model.

7. A gas leak identification system, characterized in that: include: Data acquisition module, frequency domain feature extraction module, fuzzy entropy feature extraction module, gas leakage probability prediction module and gas leakage result determination module; The data acquisition module is used to acquire real-time sound sequence data of the area to be tested; The frequency domain feature extraction module is used to extract frequency domain features of the real-time sound sequence data of the area to be tested by using a frequency domain feature extraction algorithm to obtain frequency domain features of the real-time sound sequence data of the area to be tested; The fuzzy entropy feature extraction module is used to perform fuzzy entropy calculation on the real-time sound sequence data of the area to be measured by a fuzzy entropy calculation algorithm to obtain the fuzzy entropy features of the real-time sound sequence data of the area to be measured; The gas leakage probability prediction module is used to input the frequency characteristics and fuzzy entropy characteristics of the real-time sound sequence data of the area to be tested into the trained gas leakage probability prediction model, so that the trained gas leakage probability prediction model predicts the leakage probability of the frequency characteristics and the fuzzy characteristics, and obtains the predicted value of the gas leakage probability of the area to be tested; wherein, the model training of the gas leakage probability prediction model is to use the historical sound sequence data with the actual value of the gas leakage probability of the area to be tested under different environmental conditions as the model training data for model training; The gas leakage result determination module is used to determine the gas leakage result of the area to be tested according to the predicted value of the gas leakage probability of the area to be tested.

8. A gas leakage identification system according to claim 7, characterized in that: The fuzzy entropy feature extraction module includes: a vector sequence construction unit, a vector fuzzy distance calculation unit, a vector fuzzy membership unit, a vector average fuzzy membership unit, a logarithmic average fuzzy membership unit and a fuzzy entropy feature generation unit; The vector sequence construction unit is used to construct a vector sequence for the real-time sound sequence data of the area to be tested by using a phase space reconstruction formula to obtain a sound vector sequence of the area to be tested; wherein the vector sequence satisfies the same preset similarity tolerance and different phase space dimensions; The vector fuzzy distance calculation unit is used to calculate the fuzzy distance between two vectors in the sound vector sequence by using a fuzzy distance calculation formula to obtain a plurality of vector fuzzy distances; The vector fuzzy membership unit is used to calculate the fuzzy membership of each vector fuzzy distance according to a plurality of vector fuzzy distances through a fuzzy membership function to obtain a plurality of vector fuzzy memberships; The vector average fuzzy membership unit is used to calculate the average fuzzy membership of each vector according to the plurality of vector fuzzy memberships through the average fuzzy membership function to obtain the plurality of vector average fuzzy memberships; The logarithmic average fuzzy membership unit is used to calculate the logarithmic average fuzzy membership of all vector average fuzzy memberships according to a plurality of vector average fuzzy memberships through a logarithmic average fuzzy membership calculation formula, and obtain the logarithmic average fuzzy membership of different phase space dimensions under the same similarity tolerance; The fuzzy entropy feature generating unit is used to perform fuzzy entropy calculation on the logarithmic mean fuzzy membership of different phase space dimensions under the same similarity tolerance through a fuzzy entropy calculation formula to obtain the fuzzy entropy feature of the real-time sound sequence data of the area to be measured.

9. A gas leak identification terminal device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a gas leak identification method as described in any one of claims 1 to 6 is implemented.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is executed, the device where the storage medium is located is controlled to execute a gas leakage identification method according to any one of claims 1 to 6.

Images