A gas pipeline risk intelligent monitoring method, medium, device and system

Abstract

The present application relates to a kind of gas pipeline risk intelligent monitoring method, medium, equipment and system, emergency gas shut-off valve and multidimensional information acquisition device are arranged in gas pipeline network, the real-time multimodal data of gas pipeline network is collected, and after pre-processing, the correlation characteristic quantity is extracted based on edge computing analysis method;Multi-model learning discriminator is constructed, and the running state of gas pipeline network and the probability of failure are predicted based on correlation characteristic quantity;Realize medium based on the method, equipment;System includes one or more emergency gas shut-off valves distributed in gas pipeline network, one or more multidimensional information acquisition devices, real-time multimodal data collection in gas pipeline, predict the running state of gas pipeline network and the probability of failure, and control the opening and closing of emergency gas shut-off valve, including but not limited to timely pipeline fault location, processing and early warning, avoid gas pipeline in conveying process or user gas use process to occur leakage or explosion and other accidents.

Description

Technical Field

[0001] This invention relates to the field of electronic digital data processing, and in particular to a method, medium, equipment, and system for intelligent risk monitoring of gas pipelines. Background Technology

[0002] The electromagnetic gas emergency shut-off valve is a safety emergency shut-off device for gas pipelines. It can be connected to a gas leak alarm system or to fire protection and other intelligent alarm control terminal modules. When a gas leak or other accident occurs, the alarm sounds, the gas emergency shut-off valve is energized, and the valve automatically closes, enabling rapid remote response and automatically cutting off the gas supply. This ensures the safety of gas pipeline transportation and use, and prevents fires, explosions, and other safety accidents caused by gas pipeline leaks.

[0003] However, leaks and explosions in gas pipelines and networks are often caused by a single factor, but are usually the result of multiple factors coupled together. Existing electromagnetic gas emergency shut-off valves are susceptible to environmental interference such as unstable power supply and strong external vibrations, potentially failing to cut off gas supply in a timely manner, posing a threat to gas transmission safety. Current gas pipeline safety diagnostic methods are almost all designed for single fault modes and cannot effectively and reliably trace the cause of the fault. Furthermore, electromagnetic gas emergency shut-off valves cannot automatically detect multi-dimensional characteristic quantities within the gas pipeline, including but not limited to pressure, stress, gas concentration, temperature, and flow rate; they can only activate when the electrical system fails. Therefore, for some long-term unattended gas transmission applications, it is necessary to introduce other technologies to monitor and maintain the equipment's operating status to ensure the safe operation of the gas pipeline network.

[0004] With the rapid development of artificial intelligence and Internet of Things (IoT) technologies, by integrating these technologies into the existing manufacturing and processing technology of electromagnetic gas emergency shut-off valves, real-time monitoring of characteristic quantities such as gas pressure, stress, gas concentration, temperature, and flow rate within gas pipelines can be achieved, enabling early warning judgments. This will give electromagnetic gas emergency shut-off valves more intelligent functions. Furthermore, by establishing an IoT system for gas pipeline monitoring based on electromagnetic gas emergency shut-off valves, the operational status of various monitoring nodes in the gas pipeline network can be monitored in real time, which will be of great significance in ensuring the safe delivery of gas, the safety of users' lives and property, and the maintenance of urban harmony and stability. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method, medium, equipment, and system for intelligent monitoring of gas pipeline risks.

[0006] The technical solution adopted in this invention is an intelligent risk monitoring method for gas pipelines. The method involves installing a gas emergency shut-off valve and a multi-dimensional information acquisition device within the gas pipeline network. The multi-dimensional information acquisition device collects real-time multimodal data of the gas pipeline network. After preprocessing the multimodal data, the method extracts the associated feature quantities of the gas pipeline network based on edge computing analysis. A multi-model learning discriminator is then constructed to predict the operating status of the gas pipeline network and the probability of failure based on the associated feature quantities.

[0007] Preferably, the multi-dimensional information acquisition device acquires pressure data of each monitoring node k within the gas pipeline network at time t. Stress data Gas concentration data Temperature data and gas flow data Obtain real-time multimodal data satisfy k is a positive integer, and t is a non-negative integer.

[0008] Preferably, the preprocessing of multimodal data involves denoising the obtained raw real-time multimodal data and decomposing it into intrinsic mode component data.

[0009] Preferably, decomposing intrinsic modal component data includes the following steps:

[0010] S1.1 Obtain raw data in the continuous time domain T Extracting maxima and minima, and fitting the original data using the Hermite interpolation method. The upper envelope, lower envelope, and mean coverage area of ​​the envelope. Record the original signal The intermediate signal is satisfy

[0011]

[0012] S1.2 In the continuous time domain T, if If the number of local minima and the number of zero-crossings of the data are the same, and the upper and lower envelopes of the upper and lower envelopes are both zero, then this intermediate signal... This is the first intrinsic mode component; otherwise, repeat equation (1) until the condition is met, denoted as the intrinsic mode component.

[0013] S1.3 Filter the first intrinsic mode data from the original signal, denoted as

[0014]

[0015] Will As the original signal for the first decomposition, this process is repeated n times until no intrinsic mode component functions can be decomposed from the original signal. It is decomposed into n intrinsic mode data and residual terms. Satisfying equation (3),

[0016]

[0017] S1.4 Add white noise to the intrinsic mode data to change the distribution of extreme points of the original signal and perform smoothing processing.

[0018] Preferably, S1.4 includes the following steps:

[0019] S1.4.1 in the original signal Add white noise Receive signal Satisfying equation (4),

[0020]

[0021] Where j = 1, 2, ..., m, j is the number of experiments with white noise added;

[0022] S1.4.2 The intrinsic modal components are solved according to equations (1) to (3) to obtain the o-th intrinsic modal component of the j-th test. satisfy

[0023]

[0024] in, For the i-th intrinsic mode component of the j-th test, Let N be the residual of the j-th experiment, and N be the number of experiments obtained in the j-th experiment to obtain the intrinsic modes.

[0025] S1.4.3 Calculate the mean of the eigencomponent set obtained from each decomposition, satisfying equation (6).

[0026]

[0027] Preferably, the extraction of associated feature quantities of gas pipeline networks based on edge computing analysis methods includes the following steps:

[0028] S2.1 Perform Hilbert transform on the intrinsic mode components to satisfy equation (7).

[0029]

[0030] Where ρ is the Cauchy principal quantity of the generalization integral. Let i be the i-th eigenmode component at time τ;

[0031] S2.2 Calculate the analytic function Z(T), which satisfies equation (8).

[0032]

[0033] in, Let be the complex components in the analytic function, a(T) be the amplitude function, and θ(T) be the phase function, satisfying equations (9) and (10) respectively.

[0034]

[0035]

[0036] S2.3 yields the time-domain spectral characteristics of the data set, satisfying equation (11).

[0037]

[0038] Preferably, constructing the multi-model learning discriminator includes the following steps:

[0039] S3.1 Select positive and negative samples to establish a training set (I1, y1),...,(I n ,y n ), where I n The time-domain spectral characteristics of the input, y i These are the labels for positive and negative samples, with values ​​of 0 or 1.

[0040] S3.2 Extract the correlation features of positive and negative samples using edge computing analysis methods and establish a feature vector set;

[0041] S3.3 Initialize the weights ζ corresponding to the positive and negative samples respectively. 1,i The weights are normalized.

[0042] S3.4 For each feature I i Train a classifier whose error rate is proportional to the weights ζ of the sample distribution. t Measure, ζ t Satisfying equation (13),

[0043] ζ t =∑ i ζ i |h j (I i )-y i | (13)

[0044] S3.5 Single Learning ν j (x) satisfies equation (14),

[0045]

[0046] Among them, f j For features (referring generally to the data features extracted above), θ j p is the threshold. j The bias bit is a parameter of the classifier obtained by training multiple single features in the above single learning formula. It includes the threshold, which is obtained through self-training based on the data features of each category. Here, single learning refers to learning to classify a certain type of feature. Since the data collected above is multimodal and multi-class data, we first perform single learning on one type of data, and then realize the training and learning of multi-model data.

[0047] Sort the classifiers by their error rates and select the classifier with the lowest error rate. j Its error rate is ε t Update weights When sample I i When correctly classified, e i =0, otherwise e i =1,β t Satisfying equation (15),

[0048]

[0049] S3.6 Construct a multi-model learning discriminator that satisfies equation (16).

[0050]

[0051] in,

[0052] A computer-readable storage medium storing a gas pipeline risk intelligent monitoring program thereon, which, when executed by a processor, implements the above-described gas pipeline risk intelligent monitoring method.

[0053] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the above-described intelligent risk monitoring method for gas pipelines.

[0054] A gas pipeline risk intelligent monitoring system, the system comprising:

[0055] One or more gas emergency shut-off valves are distributed within the gas pipeline network;

[0056] One or more multi-dimensional information acquisition devices are distributed within the gas pipeline network to collect real-time multimodal data of the gas pipeline network;

[0057] One control terminal, associated with a multi-dimensional information acquisition device, employs the aforementioned intelligent risk monitoring method for gas pipelines to dynamically monitor the gas pipeline network, predict its operating status and the probability of failure, and control the opening and closing of the gas emergency shut-off valve.

[0058] This invention provides a method, medium, equipment, and system for intelligent risk monitoring of gas pipelines. The method involves installing emergency gas shut-off valves and multi-dimensional information acquisition devices within the gas pipeline network to collect real-time multimodal data of the gas pipeline network. After preprocessing the multimodal data, correlation features of the gas pipeline network are extracted based on edge computing analysis. A multi-model learning discriminator is constructed to predict the operating status of the gas pipeline network and the probability of failure based on the correlation features. The method is implemented using the medium and equipment. The system includes one or more emergency gas shut-off valves and one or more multi-dimensional information acquisition devices distributed within the gas pipeline network. It achieves real-time multimodal data acquisition within the gas pipeline, predicts the operating status of the gas pipeline network and the probability of failure, and controls the opening and closing of the emergency gas shut-off valves. This includes, but is not limited to, timely fault location, handling, and early warning of pipeline network failures, preventing accidents such as leaks or explosions during gas transmission or user gas usage.

[0059] The beneficial effects of this invention are as follows:

[0060] (1) By utilizing edge computing and artificial intelligence technologies, real-time monitoring of gas in gas pipelines can be achieved, and information such as gas flow, temperature and pressure can be analyzed and processed to predict the probability of gas leakage and provide timely warnings and treatment to avoid gas accidents.

[0061] (2) Build a system, create cloud server and smart Internet of Things technology to support more monitoring devices to access and collect, analyze and transmit data, and feed the data back to the control terminal in real time to realize the remote control of the shut-off valve of the system terminal to avoid safety accidents such as fire and explosion caused by gas leakage;

[0062] (3) Realize dynamic monitoring of the operating status of gas pipelines and calculation of fault probability, realize accident early warning, which is of great significance in safety monitoring and fault assessment and prediction of gas transmission and use, and facilitates the intelligent application expansion of electromagnetic gas emergency shut-off valves. Attached Figure Description

[0063] Figure 1 This is a flowchart of the method of the present invention;

[0064] Figure 2 This is a schematic diagram of the system structure of the present invention;

[0065] Figure 3 This is a schematic diagram illustrating the application of the present invention. Detailed Implementation

[0066] The present invention will be further described in detail below with reference to embodiments, but the scope of protection of the present invention is not limited thereto.

[0067] This invention relates to an intelligent risk monitoring method for gas pipelines. The method involves installing a gas emergency shut-off valve and a multi-dimensional information acquisition device within the gas pipeline network. The multi-dimensional information acquisition device collects real-time multimodal data of the gas pipeline network. After preprocessing the multimodal data, the method extracts the associated feature quantities of the gas pipeline network based on edge computing analysis. A multi-model learning discriminator is then constructed to predict the operating status of the gas pipeline network and the probability of failure based on the associated feature quantities.

[0068] In this invention, by collecting and analyzing characteristic quantities within gas pipelines, parameters of the corresponding electromagnetic emergency shut-off valve monitoring nodes of the gas pipeline are monitored in real time, including but not limited to pressure, stress, gas concentration, temperature, and flow rate. Through artificial intelligence algorithm design, calculations and analyses are performed at the edge, and data collected by gas IoT nodes are monitored and analyzed from multiple dimensions. This enables the prediction and early warning of the probability of gas pipeline leaks and explosions, and the prediction results are fed back to the control terminal. By remotely controlling and operating the electromagnetic valve system of the pipeline nodes in advance, emergency shut-off or closure of gas valves is implemented, ensuring that gas transmission and user use are more intelligent and safer.

[0069] The method will be explained below with reference to specific content.

[0070] The multi-dimensional information acquisition device collects pressure data from each monitoring node k within the gas pipeline network at time t. Stress data Gas concentration data Temperature data and gas flow data Obtain real-time multimodal data satisfy k is a positive integer, and t is a non-negative integer.

[0071] The preprocessing of multimodal data involves denoising the obtained raw real-time multimodal data and decomposing it into intrinsic mode component data.

[0072] In this invention, by installing multi-dimensional information acquisition devices, including but not limited to pressure sensors, stress sensors, gas concentration detectors, temperature sensors, and gas flow meters, on the solenoid valve, the common data of gas pipelines can be collected, acquired, and risk and fault analysis can be achieved. The pressure sensors, stress sensors, gas concentration detectors, temperature sensors, and gas flow meters can be set separately or integrated, which is easy for those skilled in the art to understand. The integrated multi-dimensional information acquisition device can more objectively reflect the status of a certain monitoring node in the gas pipeline network.

[0073] In this invention, the pressure data of each monitoring node k at time t is used... Stress data Gas concentration data Temperature data and gas flow data The analysis enables online monitoring and real-time early warning of the operating status of gas pipelines;

[0074] Before analysis, data preprocessing is required. Since the data acquired at the monitoring nodes is time-series data, i.e., time sampling of multidimensional data, and in actual operation, the data acquired by the monitoring nodes is affected by factors such as pipe or valve wear, corrosion or deformation, and external vibration, the acquired data basically contains nonlinear and non-stationary noise. Therefore, it is necessary to first denoise the sampled data at the edge to retain the intrinsic data and then extract better data features.

[0075] In this invention, the acquired signal data is divided into three parts: noise data, intrinsic data, and time-drift data. The acquired raw data is denoised according to the time scale characteristics, and the raw data is decomposed into a set of intrinsic mode component data. The data is then transformed by Hilbert to obtain the time domain spectrum form of the data, and the intrinsic values ​​of the data are further obtained.

[0076] Specifically, decomposing intrinsic mode component data includes the following steps:

[0077] S1.1 Obtain raw data in the continuous time domain T Calculate the maxima and minima, and considering the temporal continuity of the data in the continuous time domain, fit the original data based on the Hermite interpolation method. The upper envelope, lower envelope, and mean coverage area of ​​the envelope. Let the intermediate signal of the original signal be . satisfy

[0078]

[0079] S1.2 In the continuous time domain T, if If the number of local minima and the number of zero-crossings of the data are the same, and the upper and lower envelopes of the upper and lower envelopes are both zero, then this intermediate signal... This is the first intrinsic mode component; otherwise, repeat equation (1) until the condition is met, denoted as the intrinsic mode component.

[0080] S1.3 Filter the first intrinsic mode data from the original signal, denoted as

[0081]

[0082] Will As the original signal for the first decomposition, the above operation is repeated n times until the intrinsic mode component functions can no longer be decomposed. It is decomposed into n intrinsic mode data and residual terms. Satisfying equation (3),

[0083]

[0084] S1.4 Add white noise to the intrinsic mode data to change the distribution of extreme points of the original signal and perform smoothing processing.

[0085] In this invention, because the original signal contains multiple signals with discontinuous extreme points and small frequency differences, the intrinsic mode components may exhibit "overshoot" or "undershoot," leading to mode aliasing in the original multidimensional component data. Therefore, white noise needs to be added to the intrinsic mode data to change the distribution of extreme points in the original signal and to perform smoothing processing to facilitate subsequent training. Specifically, this includes the following steps:

[0086] S1.4.1 in the original signal Add white noise Receive signal Satisfying equation (4),

[0087]

[0088] Where j = 1, 2, ..., m, and m is the number of experiments with white noise added;

[0089] S1.4.2 The intrinsic modal components are solved according to equations (1) to (3) to obtain the i-th intrinsic modal component of the j-th test. satisfy

[0090]

[0091] in, For the i-th intrinsic mode component of the j-th test, Let N be the residual of the j-th experiment, and N be the number of intrinsic modes in the j-th experiment.

[0092] S1.4.3 Calculate the mean of the eigencomponent set obtained from each decomposition, satisfying equation (6).

[0093]

[0094] After data preprocessing, the aforementioned intrinsic mode component set is subjected to Hilbert analysis to obtain the time-domain spectral characteristics of the original data. Specifically, the extraction of correlation features of the gas pipeline network based on edge computing analysis includes the following steps:

[0095] S2.1 Perform Hilbert transform on the intrinsic mode components to satisfy equation (7).

[0096]

[0097] Where ρ is the Cauchy principal quantity of the generalization integral. Let i be the i-th eigenmode component at time τ;

[0098] S2.2 Calculate the analytic function Z(T), which satisfies equation (8).

[0099]

[0100] in, Let be the complex components in the analytic function, a(T) be the amplitude function, and θ(T) be the phase function, satisfying equations (9) and (10) respectively.

[0101]

[0102] S2.3 yields the time-domain spectral characteristics of the data set, satisfying equation (11).

[0103]

[0104] Considering that the data collected on-site is multidimensional and that the features of each category are interrelated, the feature representation ability of a single feature model is not strong enough and the model learning efficiency is low. Therefore, a multi-model learning discriminator is adopted to combine multiple single models through an integration strategy, which weakens the preference on a single learning model, so that multiple single models can learn features better and more comprehensively, giving full play to the advantages of each model and avoiding the limitations of a single model.

[0105] Specifically, constructing the multi-model learning discriminator includes the following steps:

[0106] S3.1 Select positive and negative samples, typically multidimensional data collected from gas pipelines without fault warnings and multidimensional data obtained from gas pipelines with faults, respectively, to establish a training set (I1, y1),...,(I n ,y n ), where I i The time-domain spectral characteristics of the input, y i Let i be the labels for positive and negative samples, where i is 0 or 1. Generally, y i =0, 1 respectively represent that the output is a negative sample (fault) and a positive sample (normal);

[0107] S3.2 For the samples in the training set, extract the correlation features of positive and negative samples using the edge computing analysis method, and establish a feature vector set;

[0108] S3.3 Initialize the weights ζ corresponding to the positive and negative samples respectively. 1,i To be precise, it corresponds to y i =0,1, initialize weight ζ 1,i for and Where m and l are the number of negative and positive samples, respectively; the weights are normalized as shown in equation (12).

[0109]

[0110] S3.4 Train a classifier for each feature, with the classifier's error rate calculated using the weights ζ of the sample distribution. t Measure, ζ t Satisfying equation (13),

[0111] ζ t =∑ i ζ i |h j (x i )-y i | (13)

[0112] S3.5 Single Learning ν j (x) satisfies equation (14),

[0113]

[0114] Among them, f j As a characteristic, θ j p is the threshold. j This is the offset position (indicating the direction of the inequality);

[0115] Sort the classifiers by their error rates and select the classifier with the lowest error rate. t Its error rate is ε t Update weights When sample I i When correctly classified, e i =0, otherwise e i =1,β t Satisfying equation (15),

[0116]

[0117] S3.6 Construct a multi-model learning discriminator that satisfies equation (16).

[0118]

[0119] in,

[0120] The algorithm described above can be used to monitor the operational status of the gas pipeline network at the monitoring node in real time, and to provide early warning and real-time analysis of the probability of failure at that node.

[0121] The present invention also relates to a computer-readable storage medium storing a gas pipeline risk intelligent monitoring program thereon, which, when executed by a processor, implements the above-described gas pipeline risk intelligent monitoring method.

[0122] The present invention also relates to a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the above-mentioned intelligent monitoring method for gas pipeline risks.

[0123] This invention also relates to an intelligent risk monitoring system for gas pipelines, the system comprising:

[0124] One or more gas emergency shut-off valves are distributed within the gas pipeline network;

[0125] One or more multi-dimensional information acquisition devices are distributed within the gas pipeline network to collect real-time multimodal data of the gas pipeline network;

[0126] One control terminal, associated with a multi-dimensional information acquisition device, employs the aforementioned intelligent risk monitoring method for gas pipelines to dynamically monitor the gas pipeline network, predict its operating status and the probability of failure, and control the opening and closing of the gas emergency shut-off valve.

[0127] In this invention, supervised learning is performed on the multimodal feature data acquired from the emergency shut-off valve monitoring node in the gas pipeline network system to achieve online monitoring and real-time early warning of the gas pipeline network's operating status. Supervised learning is employed to learn from and model empirical data, and to predict the data, thus solving a binary classification problem in data representation. A gas pipeline risk intelligent monitoring system is constructed with the gas emergency shut-off valve as the monitoring node. 5G technology is used to remotely transmit the collected monitoring data and prediction results between the multidimensional information acquisition device and the control terminal, thereby completing the overall monitoring and control of the gas pipeline network by the gas control terminal (control terminal), preventing accidents such as leaks or explosions during gas transmission or user gas usage.

[0128] In this invention, a smart monitoring IoT system platform for gas pipeline risks can also be built during application. By collecting data at gas pipeline nodes, edge processing is performed on pressure, stress, gas concentration, flow rate, and temperature, and early warnings are submitted at the platform. The platform then feeds back to the specific gas pipeline nodes to achieve control.

[0129] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0130] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0131] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0132] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0133] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0134] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for intelligent monitoring of risks in gas pipelines, characterized in that: The method involves installing emergency gas shut-off valves and multi-dimensional information acquisition devices within the gas pipeline network. The multi-dimensional information acquisition devices collect data from each monitoring node within the gas pipeline network. exist Stress data at any time Stress data Gas concentration data Temperature data and gas flow data To obtain real-time multimodal data ,satisfy ; It is a positive integer. It is a non-negative integer; After preprocessing the multimodal data, it is decomposed into intrinsic mode component data. White noise is added to change the distribution of extreme points of the original signal, and smoothing is performed. Based on edge computing analysis, the associated feature quantities of the gas pipeline network are extracted, including the following steps: S1.1 Obtaining the continuous time domain raw data Extract the maxima and minima, and fit the original data using the Hermite interpolation method. The upper envelope, lower envelope, and mean coverage area of ​​the envelope Let the intermediate signal of the original signal be denoted as . ,satisfy (1) S1.2 in the continuous time domain Up, if If the number of local minima and the number of zero-crossings of the data are the same, and the upper and lower envelopes of the upper and lower envelopes are both zero, then this intermediate signal... This is the first intrinsic mode component; otherwise, repeat equation (1) until the condition is met, denoted as the intrinsic mode component. ; S1.3 Select the first intrinsic mode data from the original signal, denoted as... , (2) Will As the original signal for the first decomposition, repeat The process continues until the intrinsic mode component functions cannot be decomposed, and the original signal... It is decomposed into n intrinsic mode data and residual terms. Satisfying equation (3), (3) S1.4 Add white noise to the intrinsic mode data to change the distribution of extreme points of the original signal and perform smoothing processing; S2.1 Perform Hilbert transform on the intrinsic mode components to satisfy equation (7). (7) in, For the Cauchy principal quantity of the generalization integral, For a moment The first time One intrinsic mode component; S2.2 Calculate the analytic function Satisfying equation (8), (8) in, For the complex components in an analytic function. It is an amplitude function. Let be the phase function, satisfying equations (9) and (10) respectively. (9) (10) S2.3 Obtains the time-domain spectral characteristics of the data set, satisfying equation (11). (11) Constructing a multi-model learning discriminator includes the following steps: S3.1 Select positive and negative samples to establish a training set. ,in, The time-domain spectral characteristics representing the input, These are the labels for positive and negative samples, with values ​​of 0 or 1. S3.2 Extract the correlation features of positive and negative samples using edge computing analysis methods and establish a feature vector set; S3.3 Initialize the weights corresponding to the positive and negative samples respectively. The weights are normalized. S3.4 For each feature Train a classifier whose error rate is calculated using the weights of the sample distribution. measure; S3.5 Sort the classifiers by their error rates and select the classifier with the lowest error rate. Update the weights; S3.6 Construct a multi-model learning discriminator; Based on the associated feature quantities, the operating status of the gas pipeline network and the probability of failure are predicted, and the opening and closing of the gas emergency shut-off valve are controlled.

2. The intelligent risk monitoring method for gas pipelines according to claim 1, characterized in that: The preprocessing of multimodal data involves denoising the obtained raw real-time multimodal data and decomposing it into intrinsic mode component data.

3. The intelligent risk monitoring method for gas pipelines according to claim 1, characterized in that: S1.4 includes the following steps: S1.4.1 In the original signal Add white noise Receive signal Satisfying equation (4), (4) in, , The number of experiments in which white noise was added; S1.4.2 to Solving for the intrinsic modal components using equations (1) to (3) yields the ... The first trial Each intrinsic mode component ,satisfy (5) in, For the first The first trial Each intrinsic mode component For the first Residual from the second experiment For the first The number of times the intrinsic modal experiments are obtained is then determined. S1.4.3 Calculate the mean of the set of eigencomponents obtained from each decomposition, satisfying equation (6). (6)。 4. The intelligent risk monitoring method for gas pipelines according to claim 1, characterized in that: In S3.4, Satisfying equation (13), (13) Single learning in S3.5 Satisfying equation (14), (14) in, As a feature, For the threshold, This is the offset bit; Sort the classifiers by their error rates and select the classifier with the lowest error rate. Its error rate is Update weights When the sample When correctly classified, ,otherwise , Satisfying equation (15), (15) In S3.6, a multi-model learning discriminator is constructed, satisfying equation (16). (16) in, .

5. A computer-readable storage medium, characterized in that: It stores a gas pipeline risk intelligent monitoring program, which, when executed by a processor, implements the gas pipeline risk intelligent monitoring method as described in any one of claims 1 to 4.

6. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the intelligent risk monitoring method for gas pipelines as described in any one of claims 1 to 4.

7. A gas pipeline risk intelligent monitoring system, characterized in that: The system includes: One or more gas emergency shut-off valves are distributed within the gas pipeline network; One or more multi-dimensional information acquisition devices are distributed within the gas pipeline network to collect real-time multimodal data of the gas pipeline network; A control terminal, associated with a multi-dimensional information acquisition device, employs the intelligent risk monitoring method for gas pipelines as described in any one of claims 1 to 4, for dynamically monitoring the gas pipeline network, predicting the operating status of the gas pipeline network and the probability of failure, and controlling the opening and closing of the gas emergency shut-off valve.

Images