Bidirectional metering interconnected gas pipeline monitoring system and method

Abstract

The present application discloses a two-way metering interconnected gas pipeline monitoring system and method, which monitors and collects the pressure data and flow data of key positions of the pipeline through pressure sensors and flow sensors arranged along the pipeline, and introduces artificial intelligence-based data processing and analysis algorithms at the back end to perform collaborative and correlation analysis on the pressure data and flow data of these key positions of the pipeline, so as to learn and capture the dynamic interaction information between the pressure correlation characteristics and flow correlation characteristics between multiple key positions of the gas pipeline, and detect the leakage problem of the gas pipeline with this dynamic interaction information. In this way, the gas pipeline can be bidirectionally metered and monitored, thereby realizing a comprehensive analysis of the operating status of the pipeline system, and providing more reliable protection for the safe operation of the pipeline system.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to a two-way metering interconnected gas pipeline monitoring system and method. Background Art

[0002] With the acceleration of urbanization and the improvement of industrialization, gas pipeline systems, as an important part of urban infrastructure, are increasingly used in cities. However, once a gas pipeline system leaks, it may cause serious safety accidents and even endanger the safety of people's lives and property. Therefore, it is particularly important to effectively monitor and manage the gas pipeline system. Therefore, it is crucial to effectively monitor the gas pipeline system to ensure its safe operation to protect people's lives and industrial production.

[0003] Traditional gas pipeline monitoring methods mainly rely on manual inspections and regular maintenance, which have problems such as large monitoring blind spots, low monitoring frequency, and high monitoring costs, making it difficult to detect potential safety risks in a timely manner. In addition, some traditional gas pipeline monitoring systems rely on the comparison between single sensor data and thresholds to perform risk monitoring. This method is usually based on static models or rules, and a single sensor can only provide limited information and cannot fully reflect the overall operating status of the pipeline system.

[0004] Therefore, a two-way metering interconnected gas pipeline monitoring system is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a two-way metering interconnected gas pipeline monitoring system and method, which monitors and collects the pressure data and flow data of key positions of the pipeline through pressure sensors and flow sensors arranged along the pipeline, and introduces artificial intelligence-based data processing and analysis algorithms at the back end to perform collaborative and correlation analysis on the pressure data and flow data of these key positions of the pipeline, so as to learn and capture the dynamic interaction information between the pressure correlation characteristics and flow correlation characteristics between multiple key positions of the gas pipeline, and use this dynamic interaction information to detect leakage problems in the gas pipeline. In this way, the gas pipeline can be bidirectionally metered and monitored, thereby realizing a comprehensive analysis of the operating status of the pipeline system, providing more reliable protection for the safe operation of the pipeline system.

[0006] According to one aspect of the present application, a bidirectional metering interconnected gas pipeline monitoring system is provided, which includes: a pipeline monitoring data acquisition module, which is used to collect a sequence of pipeline key position pressures and a sequence of pipeline key position flows through pressure sensors and flow sensors arranged along the pipeline; a pipeline data key position correlation coding module, which is used to perform sequence coding processing on the sequence of pipeline key position pressures and the sequence of pipeline key position flows to obtain pipeline pressure position correlation feature vectors and pipeline flow position correlation feature vectors; a pipeline position data coding feature autocorrelation enhancement module, which is used to perform autocorrelation attention enhancement on the pipeline pressure position correlation feature vectors and the pipeline flow position correlation feature vectors to obtain autocorrelation enhanced pipeline pressure position correlation feature vectors and autocorrelation enhanced pipeline flow position correlation feature vectors; a gas pipeline multimodal data feature collaborative interaction module, which is used to input the autocorrelation enhanced pipeline pressure position correlation feature vectors and the autocorrelation enhanced pipeline flow position correlation feature vectors into a feature dynamic interaction module based on a priori distribution to obtain a gas pipeline multimodal collaborative interaction representation vector; A leakage detection module is used to determine a detection result based on the gas pipeline multimodal collaborative interaction representation vector, and the detection result is used to indicate whether there is a leakage; wherein the gas pipeline multimodal data feature collaborative interaction module includes: a pipeline pressure-pipeline flow position-by-position interaction unit, which is used to calculate the position-by-position division between the autocorrelation enhanced pipeline flow position-by-position correlation feature vector and the autocorrelation enhanced pipeline pressure position-by-position correlation feature vector to obtain a pipeline pressure-pipeline flow position-by-position response interaction feature vector; a response prior factor calculation unit, which is used to respectively calculate the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain a first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and a second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector; a pipeline pressure-pipeline flow responsiveness collaborative fusion unit, which is used to calculate the position-by-position sum between the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector to obtain the gas pipeline multimodal collaborative interaction representation vector.

[0007] According to another aspect of the present application, a bidirectional metering interconnected gas pipeline monitoring method is provided, which includes: collecting a sequence of pipeline key position pressures and a sequence of pipeline key position flows through pressure sensors and flow sensors arranged along the pipeline; performing sequence encoding processing on the sequence of pipeline key position pressures and the sequence of pipeline key position flows to obtain pipeline pressure position correlation feature vectors and pipeline flow position correlation feature vectors; performing autocorrelation attention enhancement on the pipeline pressure position correlation feature vectors and the pipeline flow position correlation feature vectors to obtain autocorrelation enhanced pipeline pressure position correlation feature vectors and autocorrelation enhanced pipeline flow position correlation feature vectors; inputting the autocorrelation enhanced pipeline pressure position correlation feature vectors and the autocorrelation enhanced pipeline flow position correlation feature vectors into a feature dynamic interaction module based on a priori distribution to obtain a gas pipeline multimodal collaborative interaction representation vector; based on the gas pipeline multimodal collaborative interaction representation vector, determining a detection result, the detection result Used to indicate whether there is a leak; wherein the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector are input into a feature dynamic interaction module based on a prior distribution to obtain a gas pipeline multimodal collaborative interaction representation vector, including: calculating the position-by-position division between the autocorrelation enhanced pipeline flow position correlation feature vector and the autocorrelation enhanced pipeline pressure position correlation feature vector to obtain a pipeline pressure-pipeline flow position-by-position response interaction feature vector; respectively calculating the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain a first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and a second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector; calculating the position-by-position sum between the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector to obtain the gas pipeline multimodal collaborative interaction representation vector.

[0008] Compared with the prior art, the present application provides a bidirectional metering interconnected gas pipeline monitoring system and method, which monitors and collects the pressure data and flow data of the key positions of the pipeline through the pressure sensors and flow sensors arranged along the pipeline, and introduces the data processing and analysis algorithm based on artificial intelligence at the back end to perform collaborative and correlation analysis on the pressure data and flow data of these key positions of the pipeline, so as to learn and capture the dynamic interaction information between the pressure correlation characteristics and flow correlation characteristics between multiple key positions of the gas pipeline, and detect the leakage problem of the gas pipeline with this dynamic interaction information. In this way, the gas pipeline can be bidirectionally metered and monitored, thereby realizing a comprehensive analysis of the operating status of the pipeline system, and providing more reliable protection for the safe operation of the pipeline system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 It is a block diagram of a bidirectional metering interconnected gas pipeline monitoring system according to an embodiment of the present application.

[0011] Figure 2 This is a system architecture diagram of a bidirectional metering interconnected gas pipeline monitoring system according to an embodiment of the present application.

[0012] Figure 3 It is a block diagram of the training phase of the bidirectional metering interconnected gas pipeline monitoring system according to an embodiment of the present application.

[0013] Figure 4 The present invention is a flowchart of a method for monitoring a bidirectionally metered interconnected gas pipeline according to an embodiment of the present application. DETAILED DESCRIPTION

[0014] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0015] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0016] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0017] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0018] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0019] Traditional gas pipeline monitoring methods mainly rely on manual inspections and regular maintenance, which have problems such as large monitoring blind spots, low monitoring frequency, and high monitoring costs, making it difficult to detect potential safety risks in a timely manner. In addition, some traditional gas pipeline monitoring systems rely on the comparison between a single sensor data and a threshold to perform risk monitoring. This method is usually based on static models or rules, and a single sensor can only provide limited information and cannot fully reflect the overall operating status of the pipeline system. Therefore, a two-way metering interconnected gas pipeline monitoring system is desired.

[0020] In the technical solution of the present application, a two-way metering interconnected gas pipeline monitoring system is proposed. Figure 1 It is a block diagram of a bidirectional metering interconnected gas pipeline monitoring system according to an embodiment of the present application. Figure 2 FIG. 1 is a system architecture diagram of a two-way metering interconnected gas pipeline monitoring system according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to an embodiment of the present application, a bidirectional metering interconnected gas pipeline monitoring system 300 includes: a pipeline monitoring data acquisition module 310, which is used to collect a sequence of pipeline key position pressures and a sequence of pipeline key position flows through pressure sensors and flow sensors arranged along the pipeline; a pipeline data key position correlation encoding module 320, which is used to perform sequence encoding processing on the sequence of pipeline key position pressures and the sequence of pipeline key position flows to obtain a pipeline pressure position correlation feature vector and a pipeline flow position correlation feature vector; a pipeline position data encoding feature autocorrelation enhancement module 330, which is used to encode the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector are respectively subjected to autocorrelation attention enhancement to obtain the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector; a gas pipeline multimodal data feature collaborative interaction module 340 is used to input the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector into a feature dynamic interaction module based on prior distribution to obtain a gas pipeline multimodal collaborative interaction representation vector; a leakage detection module 350 is used to determine the detection result based on the gas pipeline multimodal collaborative interaction representation vector, and the detection result is used to indicate whether there is a leakage.

[0021] In particular, the pipeline monitoring data acquisition module 310 and the pipeline data key position correlation encoding module 320 are used to collect the sequence of pipeline key position pressure and the sequence of pipeline key position flow through the pressure sensor and flow sensor arranged along the pipeline; and the sequence of pipeline key position pressure and the sequence of pipeline key position flow are sequence-encoded to obtain the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector. In a specific example of the present application, the sequence of pipeline key position pressure and the sequence of pipeline key position flow are input into the sequence encoder based on the Bi-LSTM model to obtain the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector. Considering that the pipeline key position pressure and flow data collected by the pressure sensor and flow sensor can reflect the pressure and flow numerical information at different gas pipeline key positions, the pressure or flow data at different key positions not only show direct correlation, but also may hide deeper correlation patterns, which may be related to factors such as the physical properties, materials, service life, and maintenance status of the pipeline. That is to say, the pressure or flow data at these different key positions have data correlation relationships and hidden correlation patterns based on the parameter samples. This correlation information can reflect the pressure distribution or flow distribution of the gas pipeline, which helps to analyze the evolution of pressure and flow during transmission in the pipeline, so as to better detect whether there is a leakage problem. For example, if the pressure at a certain position suddenly drops, it may indicate a leakage or other problem. Based on this, in order to better understand the implicit correlation relationship and pattern characteristics of the pressure or flow between different key positions of the pipeline, in the technical solution of the present application, the sequence of the pipeline key position pressure and the sequence of the pipeline key position flow are further input into the sequence encoder based on the Bi-LSTM model for encoding, so as to extract the correlation characteristics and pattern information of the pressure and flow in the gas pipeline at each key position, so as to obtain the correlation feature vector between the pipeline pressure positions and the correlation feature vector between the pipeline flow positions, which is conducive to better identifying abnormal conditions of the gas pipeline, such as leakage events.

[0022] In particular, the pipeline position data encoding feature autocorrelation enhancement module 330 is used to perform autocorrelation attention enhancement on the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector to obtain the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector. In a specific example of the present application, the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector are respectively input into the autocorrelation attention module based on the Transformer structure to obtain the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector. Since the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector respectively contain the implicit correlation patterns and features of the pipeline pressure and pipeline flow at each key position of the pipeline, the encoding processing method of the Bi-LSTM model is not sensitive to the long-distance global dependency correlation relationship. In the pipeline monitoring system, there may be complex interactions between the pressure and flow data at different positions, and more global position data dependency correlation features are required to achieve more accurate detection. Based on this, in order to further integrate the contextual information between these different positions, so as to enhance the correlation characteristics of pipeline pressure or pipeline flow between different key positions, and thus extract a richer and more comprehensive pipeline state representation, in the technical solution of the present application, the correlation feature vector between the pipeline pressure positions and the correlation feature vector between the pipeline flow positions are further input into the autocorrelation attention module based on the Transformer structure to obtain the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector. Through the processing of the autocorrelation attention module based on the Transformer structure, the inherent autocorrelation connection and global features in the correlation features between pipeline pressure at different positions and the correlation features between pipeline flow at different positions can be learned and captured, and the expression of these inherent autocorrelation global features can be enhanced, so that the correlation feature vector between pipeline pressure positions and the correlation feature vector between pipeline flow positions are more representative and distinguishable. This helps the system to more comprehensively analyze the operating status of the pipeline system, so as to better identify and distinguish the normal operating status and potential abnormal conditions of the gas pipeline.

[0023] In particular, the gas pipeline multimodal data feature collaborative interaction module 340 is used to input the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector into the feature dynamic interaction module based on the prior distribution to obtain the gas pipeline multimodal collaborative interaction representation vector. It should be understood that since the correlation features of pipeline pressure at different positions and the correlation features of pipeline flow at different positions after autocorrelation enhancement have dynamic interaction and hidden correlation features, that is, the pipeline pressure at different positions and the pipeline flow at different positions are mutually influenced and acted upon. If there is a gas pipeline leakage problem, there will be different correlation modes and abnormal action mechanisms, which will help to detect pipeline leakage more accurately. Based on this, in order to be able to dynamically interact and collaboratively understand the correlation features between pipeline pressure positions and pipeline flow positions at different key positions, in the technical solution of the present application, the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector are further input into the feature dynamic interaction module based on the prior distribution to obtain the gas pipeline multimodal collaborative interaction representation vector. Through the processing of the feature dynamic interaction module based on prior distribution, the pipeline pressure position correlation feature and pipeline flow position correlation feature can be used as prior information to automatically learn and capture the interaction and response mechanism between pipeline pressure and flow at different positions, and according to the interactive synergy, the pipeline pressure position correlation feature and pipeline flow position correlation feature are adaptively weighted in the process of fusion of the two features to adjust their contribution to information sharing in the fusion process. This helps to improve the system's ability to understand and judge the operating status of the pipeline system, and further improve the accuracy and stability of the monitoring system.

[0024] In an embodiment of the present application, the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector are input into a feature dynamic interaction module based on a prior distribution to obtain a gas pipeline multimodal collaborative interaction representation vector, including: first calculating the position-by-position division between the autocorrelation enhanced pipeline flow position correlation feature vector and the autocorrelation enhanced pipeline pressure position correlation feature vector to obtain a pipeline pressure-pipeline flow position-by-position response interaction feature vector; then, respectively calculating the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain a first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and a second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector; then calculating the position-by-position sum between the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector to obtain the gas pipeline multimodal collaborative interaction representation vector.

[0025] In a specific example, the process of respectively calculating the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector includes: multiplying the pipeline pressure-pipeline flow position-by-position response interaction feature vector by a first predetermined weight hyperparameter position by position to obtain a first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector; using each position feature value in the first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector as an exponent of a natural constant to calculate an exponential function value based on the natural constant according to the position to obtain a first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector; multiplying the first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector by a first predetermined weight hyperparameter position by position to obtain a first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector. The interactive class support feature vector is multiplied by the first Gaussian distribution random number function value to obtain the first pipeline pressure-pipeline flow position-by-position response interactive prior feature vector; the pipeline pressure-pipeline flow position-by-position response interactive feature vector is multiplied by the second predetermined weight hyperparameter position by position to obtain the second weight-modulated pipeline pressure-pipeline flow position-by-position response interactive feature vector; the position feature values ​​of each position in the second weight-modulated pipeline pressure-pipeline flow position-by-position response interactive feature vector are used as the exponent of the natural constant to calculate the exponential function value based on the natural constant according to the position to obtain the second weight-modulated pipeline pressure-pipeline flow position-by-position response interactive class support feature vector; the second weight-modulated pipeline pressure-pipeline flow position-by-position response interactive class support feature vector is multiplied by the second Gaussian distribution random number function value to obtain the second pipeline pressure-pipeline flow position-by-position response interactive prior feature vector. Wherein, the first Gaussian distribution random number function value and the second Gaussian distribution random number function value are both generated by a Gaussian distribution random number function with a mean of 0 and a variance of 1.

[0026] In summary, in the above embodiment, the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector are input into the feature dynamic interaction module based on prior distribution to obtain the gas pipeline multimodal collaborative interaction representation vector, including: inputting the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector into the feature dynamic interaction module based on prior distribution and processing them with the following feature dynamic interaction formula to obtain the gas pipeline multimodal collaborative interaction representation vector; wherein the feature dynamic interaction formula is: ;in, is the correlation feature vector between the autocorrelation enhanced pipeline pressure positions, is the correlation feature vector between the autocorrelation-enhanced pipeline flow positions, and is the predetermined weight hyperparameter, is the pipeline pressure-pipeline flow location-by-location response interaction feature vector, and The hyperparameters of the Gaussian distribution function coefficients are generated by a Gaussian distribution random number function with a mean of 0 and a variance of 1. is vector addition, is a representation vector of the multimodal collaborative interaction of the gas pipeline.

[0027] In particular, the leakage detection module 350 is used to determine the detection result based on the multimodal collaborative interaction representation vector of the gas pipeline, and the detection result is used to indicate whether there is a leak. In a specific example of the present application, the multimodal collaborative interaction representation vector of the gas pipeline is input into a classifier-based leakage detector to obtain a detection result, and the detection result is used to indicate whether there is a leak. That is, the multimodal collaborative interaction representation features of the gas pipeline are used to perform classification processing to detect leakage problems in the gas pipeline. In this way, bidirectional metering and monitoring of the gas pipeline can be performed, thereby achieving a comprehensive analysis of the operating status of the pipeline system and providing more reliable protection for the safe operation of the pipeline system.

[0028] Here, for the local time domain short-range-long-range bidirectional temporal correlation features of the pipeline key position pressure and the pipeline key position flow respectively expressed by the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector based on local time domain attention enhancement, during the dynamic interaction of features based on the prior distribution, the difference in feature dynamic interaction weights caused by the difference in temporal feature distribution will lead to the multimodal temporal feature aggregation offset of the obtained multimodal collaborative interaction representation vector of the gas pipeline. Therefore, it is expected to further improve the comprehensibility of the multimodal temporal feature aggregation regression of the multimodal collaborative interaction representation vector of the gas pipeline, thereby improving the accuracy of the detection results obtained by the leakage detector based on the classifier when it is input.

[0029] In a preferred embodiment, the gas pipeline multimodal collaborative interaction representation vector is input into a classifier-based leakage detector to obtain a detection result, including: subtracting the zero norm of the gas pipeline multimodal collaborative interaction representation vector from the length of the gas pipeline multimodal collaborative interaction representation vector to obtain the gas pipeline multimodal collaborative interaction isolated zero value; calculating the logarithm of the square of the gas pipeline multimodal collaborative interaction isolated zero value and the sum of the gas pipeline multimodal collaborative interaction isolated zero value to obtain the gas pipeline multimodal collaborative interaction superposition information value; calculating a power function with each eigenvalue of the gas pipeline multimodal collaborative interaction representation vector as the base, and the difference between the gas pipeline multimodal collaborative interaction isolated zero value and one as the exponential, and performing a point multiplication with the gas pipeline multimodal collaborative interaction superposition information value to obtain the first gas pipeline. A multimodal collaborative interaction intermediate vector; performing a dot product of the multimodal collaborative interaction representation vector of the gas pipeline and the difference of the isolated zero value of the multimodal collaborative interaction of the gas pipeline minus one, and then performing a dot product of the multimodal collaborative interaction representation vector of the gas pipeline and the reciprocal of the isolated zero value of the multimodal collaborative interaction of the gas pipeline to obtain a second multimodal collaborative interaction intermediate vector of the gas pipeline; calculating an exponential function with a natural constant as the base, in which each eigenvalue of the second multimodal collaborative interaction intermediate vector of the gas pipeline is an exponential function to obtain a third multimodal collaborative interaction intermediate vector of the gas pipeline; performing a dot addition of the first multimodal collaborative interaction intermediate vector of the gas pipeline and the third multimodal collaborative interaction intermediate vector of the gas pipeline to obtain an optimized multimodal collaborative interaction representation vector of the gas pipeline; inputting the optimized multimodal collaborative interaction representation vector of the gas pipeline into a classifier-based leakage detector to obtain a detection result.

[0030] That is, the multimodal collaborative interaction representation vector of the gas pipeline is, for example, recorded as The optimization process is expressed as: ;in, represents the multimodal collaborative interaction representation vector of the gas pipeline, represents the zero norm of a vector, represents the length of the gas pipeline multimodal collaborative interaction representation vector, represents the isolated zero value of the multi-modal collaborative interaction of the gas pipeline, It is the point multiplication by position. is added by position, represents a power function whose exponent is the difference between the isolated zero value of the multimodal cooperative interaction of the gas pipeline and one, Representation vector of optimized multimodal collaborative interaction of gas pipeline.

[0031] In this way, the high-dimensional feature manifold of the multimodal collaborative interaction representation vector of the gas pipeline is represented by the feature value of the feature set as the vector field of the aggregation dimension, and the superposition value of the vector field of the multimodal collaborative interaction representation vector of the gas pipeline is used as the order information to fix the local position of the feature value of its feature set, and add the bias of the reversibility of the feature regression distribution field of the multimodal collaborative interaction representation vector of the gas pipeline as a reward to achieve the mapping target tracking of the regression distribution of the multimodal collaborative interaction representation vector of the gas pipeline for the feature value position, so that the feature set of the multimodal collaborative interaction representation vector of the gas pipeline perceives the mapping migration to the aggregate distribution, thereby improving the accuracy of the detection result obtained by the multimodal collaborative interaction representation vector of the gas pipeline inputting the leakage detector based on the classifier by improving the multimodal time series feature aggregation regression comprehensibility of the multimodal collaborative interaction representation vector of the gas pipeline. In this way, more accurate bidirectional metering and monitoring of the gas pipeline can be carried out, so as to detect the leakage problem of the gas pipeline in a timely and effective manner, and provide more reliable guarantee for the safe operation of the pipeline system.

[0032] It should be understood that before using the above-mentioned neural network model for inference, it is necessary to train the sequence encoder based on the Bi-LSTM model, the autocorrelation attention module based on the Transformer structure, the feature dynamic interaction module based on the prior distribution, and the leakage detector based on the classifier. That is to say, according to the bidirectional metering interconnected gas pipeline monitoring system of the present application, it also includes a training phase for training the sequence encoder based on the Bi-LSTM model, the autocorrelation attention module based on the Transformer structure, the feature dynamic interaction module based on the prior distribution, and the leakage detector based on the classifier.

[0033] Figure 3 FIG. 1 is a block diagram of the training phase of the bidirectional metering interconnected gas pipeline monitoring system according to an embodiment of the present application. Figure 3As shown, according to the bidirectional metering interconnected gas pipeline monitoring system 300 of the embodiment of the present application, it includes: a training module 400, including: a training data acquisition unit 410, used to acquire training data, wherein the training data includes a sequence of training pipeline key position pressures and a sequence of training pipeline key position flows collected by pressure sensors and flow sensors arranged along the pipeline; a training data key position correlation encoding unit 420, used to input the sequence of training pipeline key position pressures and the sequence of training pipeline key position flows into a sequence encoder based on a Bi-LSTM model to obtain a training pipeline pressure position correlation feature vector and a training pipeline flow position correlation feature vector; a training data encoding feature autocorrelation enhancement unit 430, used to input the training pipeline pressure position correlation feature vector and the training pipeline flow position correlation feature vector respectively into an autocorrelation attention module based on a Transformer structure to obtain a training autocorrelation enhanced pipeline pressure position correlation feature vector and a training autocorrelation enhanced pipeline flow position correlation feature vector; a training multimodal data feature collaborative interaction unit 440 , used to input the training autocorrelation enhanced pipeline pressure position correlation feature vector and the training autocorrelation enhanced pipeline flow position correlation feature vector into a feature dynamic interaction module based on prior distribution to obtain a training gas pipeline multimodal collaborative interaction representation vector; an optimization unit 450, used to optimize the training gas pipeline multimodal collaborative interaction representation vector to obtain an optimized training gas pipeline multimodal collaborative interaction representation vector; a training detection result generation unit 460, used to pass the optimized training gas pipeline multimodal collaborative interaction representation vector through a classifier-based leakage detector to obtain a training detection result; a cross entropy loss calculation unit 470, used to calculate the cross entropy loss function value between the training detection result and the real detection result to obtain the detection result information cross entropy loss function value; a training unit 480, used to train the sequence encoder based on the Bi-LSTM model, the autocorrelation attention module based on the Transformer structure, the feature dynamic interaction module based on prior distribution and the leakage detector based on the classifier based on the detection result information cross entropy loss function value.

[0034] As described above, the two-way metering interconnected gas pipeline monitoring system 300 according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a two-way metering interconnected gas pipeline monitoring algorithm. In one possible implementation, the two-way metering interconnected gas pipeline monitoring system 300 according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the two-way metering interconnected gas pipeline monitoring system 300 can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the two-way metering interconnected gas pipeline monitoring system 300 can also be one of the many hardware modules of the wireless terminal.

[0035] Alternatively, in another example, the bidirectional metering interconnected gas pipeline monitoring system 300 and the wireless terminal may also be separate devices, and the bidirectional metering interconnected gas pipeline monitoring system 300 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0036] Furthermore, a two-way metering interconnected gas pipeline monitoring method is also provided.

[0037] Figure 4 FIG. 1 is a flow chart of a method for monitoring a bidirectionally metered interconnected gas pipeline according to an embodiment of the present application. Figure 4 As shown, according to the embodiment of the present application, the bidirectional metering interconnected gas pipeline monitoring method includes the following steps: S1, a sequence of pipeline key position pressures and a sequence of pipeline key position flows collected by pressure sensors and flow sensors arranged along the pipeline; S2, sequence encoding processing of the sequence of pipeline key position pressures and the sequence of pipeline key position flows to obtain pipeline pressure position correlation feature vectors and pipeline flow position correlation feature vectors; S3, autocorrelation attention enhancement of the pipeline pressure position correlation feature vectors and the pipeline flow position correlation feature vectors to obtain autocorrelation enhanced pipeline pressure position correlation feature vectors and autocorrelation enhanced pipeline flow position correlation feature vectors; S4, inputting the autocorrelation enhanced pipeline pressure position correlation feature vectors and the autocorrelation enhanced pipeline flow position correlation feature vectors into a feature dynamic interaction module based on prior distribution to obtain a gas pipeline multimodal collaborative interaction representation vector; S5, determining a detection result based on the gas pipeline multimodal collaborative interaction representation vector, and the detection result is used to indicate whether there is a leak.

[0038] In summary, the bidirectional metering interconnected gas pipeline monitoring method according to the embodiment of the present application is explained, which collects the pressure data and flow data of the key positions of the pipeline through the pressure sensors and flow sensors arranged along the pipeline, and introduces the data processing and analysis algorithm based on artificial intelligence at the back end to perform collaborative and correlation analysis on the pressure data and flow data of these key positions of the pipeline, so as to learn and capture the dynamic interaction information between the pressure correlation characteristics and flow correlation characteristics between multiple key positions of the gas pipeline, and detect the leakage problem of the gas pipeline with this dynamic interaction information. In this way, the gas pipeline can be bidirectionally metered and monitored, thereby realizing a comprehensive analysis of the operating status of the pipeline system, and providing more reliable protection for the safe operation of the pipeline system.

[0039] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A two-way metering interconnected gas pipeline monitoring system, characterized in that: include: A pipeline monitoring data acquisition module, used to collect the sequence of pipeline key position pressures and the sequence of pipeline key position flows through pressure sensors and flow sensors arranged along the pipeline; a pipeline data key position correlation coding module, used to perform sequence coding processing on the sequence of pipeline key position pressures and the sequence of pipeline key position flows to obtain pipeline pressure position correlation feature vectors and pipeline flow position correlation feature vectors; a pipeline position data coding feature autocorrelation enhancement module, used to perform autocorrelation attention enhancement on the pipeline pressure position correlation feature vectors and the pipeline flow position correlation feature vectors to obtain autocorrelation enhanced pipeline pressure position correlation feature vectors and autocorrelation enhanced pipeline flow position correlation feature vectors; a gas pipeline multimodal data feature collaborative interaction module, used to input the autocorrelation enhanced pipeline pressure position correlation feature vectors and the autocorrelation enhanced pipeline flow position correlation feature vectors into a feature dynamic interaction module based on a priori distribution to obtain a gas pipeline multimodal collaborative interaction representation vector; a leak detection module, used to perform autocorrelation enhancement on the pipeline pressure position correlation feature vectors and the pipeline flow position correlation feature vectors based on the gas pipeline. A multimodal collaborative interaction representation vector is used to determine the detection result, and the detection result is used to indicate whether there is a leak; wherein the gas pipeline multimodal data feature collaborative interaction module includes: a pipeline pressure-pipeline flow position-by-position interaction unit, which is used to calculate the position-by-position division between the autocorrelation enhanced pipeline flow position-to-position correlation feature vector and the autocorrelation enhanced pipeline pressure position-to-position correlation feature vector to obtain a pipeline pressure-pipeline flow position-by-position response interaction feature vector; a response prior factor calculation unit, which is used to respectively calculate the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain a first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and a second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector; a pipeline pressure-pipeline flow responsiveness collaborative fusion unit, which is used to calculate the position-by-position sum between the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector to obtain the gas pipeline multimodal collaborative interaction representation vector.

2. The two-way metering interconnected gas pipeline monitoring system according to claim 1 is characterized in that: The pipeline data key position correlation encoding module is used to: input the sequence of the pipeline key position pressure and the sequence of the pipeline key position flow into a sequence encoder based on the Bi-LSTM model to obtain the pipeline pressure position correlation feature vector and the pipeline flow position correlation feature vector.

3. The two-way metering interconnected gas pipeline monitoring system according to claim 2 is characterized in that: The pipeline position data encoding feature autocorrelation enhancement module is used to: input the pipeline pressure position correlation feature vector and the pipeline flow rate position correlation feature vector into the autocorrelation attention module based on the Transformer structure respectively to obtain the autocorrelation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow rate position correlation feature vector.

4. The two-way metering interconnected gas pipeline monitoring system according to claim 3 is characterized in that: The response prior factor calculation unit is used to: multiply the pipeline pressure-pipeline flow position-by-position response interaction feature vector by a first predetermined weight hyperparameter position by position to obtain a first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector; use each position feature value in the first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector as an exponent of a natural constant to calculate an exponential function value based on the position with the natural constant as the base to obtain a first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector; multiply the first weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector by a first Gaussian distribution random number function value to obtain the first pipeline pressure-pipeline flow position-by-position response interaction class support feature vector. Mutual prior feature vector; multiplying the pipeline pressure-pipeline flow position-by-position response interaction feature vector by a second predetermined weight hyperparameter position by position to obtain a second weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector; using each position feature value in the second weight-modulated pipeline pressure-pipeline flow position-by-position response interaction feature vector as an exponent of a natural constant to calculate an exponential function value based on the natural constant by position to obtain a second weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector; multiplying the second weight-modulated pipeline pressure-pipeline flow position-by-position response interaction class support feature vector by a second Gaussian distribution random number function value to obtain the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector.

5. The two-way metering interconnected gas pipeline monitoring system according to claim 4 is characterized in that: The first Gaussian distribution random number function value and the second Gaussian distribution random number function value are both generated by a Gaussian distribution random number function with a mean of 0 and a variance of 1.

6. The two-way metering interconnected gas pipeline monitoring system according to claim 5 is characterized in that: The leakage detection module is used to: input the gas pipeline multimodal collaborative interaction representation vector into a classifier-based leakage detector to obtain a detection result, and the detection result is used to indicate whether there is a leakage.

7. The two-way metering interconnected gas pipeline monitoring system according to claim 6 is characterized in that: It also includes a training module for training the sequence encoder based on the Bi-LSTM model, the autocorrelation attention module based on the Transformer structure, the feature dynamic interaction module based on the prior distribution, and the leakage detector based on the classifier.

8. The two-way metering interconnected gas pipeline monitoring system according to claim 7 is characterized in that: The training module includes: a training data acquisition unit, which is used to acquire training data, wherein the training data includes a sequence of pressures at key positions of the training pipeline and a sequence of flows at key positions of the training pipeline collected by pressure sensors and flow sensors arranged along the pipeline; a training data key position correlation encoding unit, which is used to input the sequence of pressures at key positions of the training pipeline and the sequence of flows at key positions of the training pipeline into a sequence encoder based on a Bi-LSTM model to obtain a training pipeline pressure position correlation feature vector and a training pipeline flow position correlation feature vector; a training data encoding feature autocorrelation enhancement unit, which is used to input the training pipeline pressure position correlation feature vector and the training pipeline flow position correlation feature vector into an autocorrelation attention module based on a Transformer structure respectively to obtain a training autocorrelation enhanced pipeline pressure position correlation feature vector and a training autocorrelation enhanced pipeline flow position correlation feature vector; a training multimodal data feature collaborative interaction unit, which is used to input the training autocorrelation enhanced pipeline pressure position correlation feature vector into a sequence encoder based on a Bi-LSTM model to obtain a training pipeline pressure position correlation feature vector and a training pipeline flow position correlation feature vector; The position correlation feature vector and the training autocorrelation enhanced pipeline flow position correlation feature vector are input into a feature dynamic interaction module based on a prior distribution to obtain a training gas pipeline multimodal collaborative interaction representation vector; an optimization unit is used to optimize the training gas pipeline multimodal collaborative interaction representation vector to obtain an optimized training gas pipeline multimodal collaborative interaction representation vector; a training detection result generation unit is used to pass the optimized training gas pipeline multimodal collaborative interaction representation vector through a classifier-based leakage detector to obtain a training detection result; a cross entropy loss calculation unit is used to calculate the cross entropy loss function value between the training detection result and the real detection result to obtain the detection result information cross entropy loss function value; a training unit is used to train the sequence encoder based on the Bi-LSTM model, the autocorrelation attention module based on the Transformer structure, the feature dynamic interaction module based on the prior distribution and the leakage detector based on the classifier based on the detection result information cross entropy loss function value.

9. A two-way metering interconnected gas pipeline monitoring method, characterized in that: include: A sequence of pressures at key positions of the pipeline and a sequence of flows at key positions of the pipeline are collected by pressure sensors and flow sensors arranged along the pipeline; the sequence of pressures at key positions of the pipeline and the sequence of flows at key positions of the pipeline are subjected to sequence encoding processing to obtain a correlation feature vector between pipeline pressure positions and a correlation feature vector between pipeline flow positions; the correlation feature vector between pipeline pressure positions and the correlation feature vector between pipeline flow positions are subjected to autocorrelation attention enhancement to obtain a self-correlation enhanced correlation feature vector between pipeline pressure positions and a self-correlation enhanced correlation feature vector between pipeline flow positions; the self-correlation enhanced correlation feature vector between pipeline pressure positions and the self-correlation enhanced correlation feature vector between pipeline flow positions are input into a feature dynamic interaction module based on a priori distribution to obtain a multimodal collaborative interaction representation vector of the gas pipeline; based on the multimodal collaborative interaction representation vector of the gas pipeline, a detection result is determined, and the detection result is used to indicate whether there is a leak; wherein ... The correlation enhanced pipeline pressure position correlation feature vector and the autocorrelation enhanced pipeline flow position correlation feature vector are input into a feature dynamic interaction module based on prior distribution to obtain a gas pipeline multimodal collaborative interaction representation vector, including: calculating the position-by-position division between the autocorrelation enhanced pipeline flow position correlation feature vector and the autocorrelation enhanced pipeline pressure position correlation feature vector to obtain a pipeline pressure-pipeline flow position-by-position response interaction feature vector; respectively calculating the first prior factor and the second prior factor of the pipeline pressure-pipeline flow position-by-position response interaction feature vector to obtain a first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and a second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector; calculating the position-by-position sum between the first pipeline pressure-pipeline flow position-by-position response interaction prior feature vector and the second pipeline pressure-pipeline flow position-by-position response interaction prior feature vector to obtain the gas pipeline multimodal collaborative interaction representation vector.

Images