Intelligent gas operation management system and method based on AI technology

Abstract

The present application relates to the field of gas operation management, and specifically discloses an intelligent gas operation management system and method based on Ai technology, which first obtains the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas equipment parameter collector, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains the classification result through the classifier to determine whether to issue an abnormal gas energy consumption warning, thereby optimizing energy efficiency, improving the effective utilization rate of energy, effectively controlling costs, and avoiding unnecessary expenses caused by energy waste.

Description

Technical Field

[0001] The present application relates to the field of gas operation management, and more specifically, to an intelligent gas operation management system and method based on Ai technology. Background Art

[0002] With the continuous expansion of gas companies' pipeline networks and the increasing number of users, traditional management methods can no longer meet the needs of gas companies. The information island problem is particularly serious between professional systems, which increases the difficulty of information sharing and business collaboration between departments. With the advancement of Internet technology, combining information tools with modern management methods and applying them to gas production and operation management has become a common practice among gas companies.

[0003] Although modern gas utilization can achieve high efficiency and energy saving, gas energy consumption monitoring in the factory area still cannot achieve dynamic tracking and real-time statistics. In addition, the energy consumption data of each gas-consuming unit needs to be collected and recorded one by one, and the collected data information needs to be manually evaluated for gas energy efficiency, audited and publicized. This method not only fails to achieve resource optimization of each energy-consuming unit, but also makes it difficult to discover and solve the problem of gas energy waste in a timely manner.

[0004] Therefore, an intelligent gas operation management system and method based on Ai technology is desired. Summary of the invention

[0005] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides an intelligent gas operation management system and method based on Ai technology, which first obtains the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas consumption parameter data collected by the gas equipment parameter collector, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains the classification result through the classifier to determine whether to issue an abnormal gas energy consumption warning, thereby optimizing energy efficiency, improving the effective utilization rate of energy, effectively controlling costs, and avoiding unnecessary expenses caused by energy waste.

[0006] According to one aspect of the present application, there is provided an intelligent gas operation and management system based on Ai technology, which includes:

[0007] The gas operation management data acquisition module is used to obtain the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector;

[0008] A gas operation management data extraction module, used to extract a global correlation feature vector of gas supply pipe gas parameters and a dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector;

[0009] The gas energy consumption abnormal warning judgment module is used to judge whether to issue a gas energy consumption abnormal warning based on the global correlation feature vector of the gas supply pipe gas parameters and the semantic feature vector of the dynamic gas consumption parameters.

[0010] According to another aspect of the present application, there is provided an intelligent gas operation management method based on Ai technology, which includes:

[0011] Obtaining gas pressure values ​​at multiple predetermined time points collected by a gas supply pipe pressure sensor, gas flow values ​​at multiple predetermined time points collected by a gas supply pipe flow sensor, and dynamic gas usage parameter data collected by a gas usage equipment parameter collector;

[0012] Extracting a gas supply pipe gas parameter global correlation feature vector and a dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector;

[0013] Based on the global correlation feature vector of the gas supply pipe gas parameters and the semantic feature vector of the dynamic gas usage parameters, it is determined whether to issue an abnormal gas energy consumption warning.

[0014] Compared with the prior art, the present application provides an intelligent gas operation management system and method based on Ai technology, which first obtains the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas consumption parameter data collected by the gas equipment parameter collector, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains the classification result through the classifier to determine whether to issue an abnormal gas energy consumption warning, thereby optimizing energy efficiency, improving the effective utilization rate of energy, effectively controlling costs, and avoiding unnecessary expenses caused by energy waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] 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.

[0016] Figure 1 This is a block diagram of an intelligent gas operation and management system based on Ai technology according to an embodiment of the present application.

[0017] Figure 2 It is a block diagram of a gas operation management data extraction module in an intelligent gas operation management system based on Ai technology according to an embodiment of the present application.

[0018] Figure 3 This is a block diagram of a gas energy consumption abnormal warning judgment module in an intelligent gas operation and management system based on Ai technology according to an embodiment of the present application.

[0019] Figure 4 This is a flowchart of an intelligent gas operation management method based on Ai technology according to an embodiment of the present application.

[0020] Figure 5 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] 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.

[0022] Figure 1 : is a block diagram of an intelligent gas operation and management system based on AI technology according to an embodiment of the present application. Figure 1As shown, according to the embodiment of the present application, the intelligent gas operation and management system 100 based on Ai technology includes: a gas operation and management data acquisition module 110, which is used to obtain the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas equipment parameter collector; a gas operation and management data extraction module 120, which is used to extract the global correlation feature vector of the gas parameter of the gas supply pipe and the dynamic gas usage parameter data collected by the gas equipment parameter collector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data; a gas energy consumption abnormal warning judgment module 130, which is used to judge whether to issue a gas energy consumption abnormal warning based on the global correlation feature vector of the gas parameter of the gas supply pipe and the dynamic gas usage parameter semantic feature vector.

[0023] In the above-mentioned intelligent gas operation and management system 100 based on Ai technology, the gas operation and management data acquisition module 110 is used to obtain the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas consumption parameter data collected by the gas equipment parameter collector. It should be understood that with the rapid expansion of the gas enterprise pipeline network and user scale, the traditional management method has become incompatible with current needs. There are serious information island problems between various professional systems, which leads to huge challenges in information sharing and business collaboration between departments. With the advancement of Internet technology, gas companies generally recognize the importance of combining information tools with modern management methods to optimize production and operation management. Although modern gas utilization technology can achieve high efficiency and energy saving, there are still problems in dynamic tracking and real-time statistics in the gas energy consumption monitoring within the factory area. The energy consumption data of each gas energy-consuming unit needs to be collected and recorded separately, and then manual energy efficiency evaluation, energy audit and publicity are required. This method not only fails to effectively optimize the resource utilization of each unit, but also makes it difficult to timely discover and solve the problem of gas energy waste. Therefore, in the technical solution of the present application, by obtaining the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas consumption parameter data collected by the gas equipment parameter collector, and combining with deep learning technology, by identifying abnormal gas energy consumption and issuing early warnings, energy efficiency is optimized and the effective utilization rate of energy is improved, thereby effectively controlling costs and avoiding unnecessary expenses due to energy waste.

[0024] Specifically, the gas pressure value collected by the gas supply pipe pressure sensor reflects the pressure state inside the pipeline. These data are crucial to the stability of gas supply. Through the collection of multiple predetermined time points, the system can monitor the changing trend of pipeline pressure in real time, so as to adjust the gas supply in time and ensure the stability and safety of supply. In addition, the gas pressure data is also used to analyze the operating status of the pipeline, such as whether there is leakage or other abnormal conditions, and to provide a reference for future operation planning. The gas flow value collected by the gas supply pipe flow sensor reflects the change in the actual gas supply. These data are not only used for metering and settlement, but also indicate the load of the pipeline. By analyzing the flow data at multiple time points, the system can identify the peak and trough periods of the gas supply pipeline, thereby optimizing the gas supply strategy and improving the gas supply efficiency. This is an important basis for cost control and resource allocation for gas companies. The dynamic gas parameter data collected by the gas equipment parameter collector covers the real-time operation of the gas equipment. These parameters may include information such as the working status, energy consumption, and operating efficiency of the equipment. By regularly collecting and analyzing this data, the system can identify abnormal operation or signs of failure of the equipment, and perform preventive maintenance in advance to avoid adverse effects of equipment damage or shutdown on the gas supply system. The intelligent gas operation management system obtains the pressure, flow and dynamic gas usage parameter data of the gas supply pipe, realizes comprehensive monitoring and accurate analysis of the operating status of the gas system, and can respond to potential problems in a timely manner to ensure the stability, safety and economy of gas supply.

[0025] In the above-mentioned intelligent gas operation and management system 100 based on Ai technology, the gas operation management data extraction module 120 is used to extract the global correlation feature vector of the gas parameters of the gas supply pipe and the dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas equipment parameter collector. It should be understood that by integrating the global correlation feature vector of the gas parameters of the gas supply pipe and the dynamic gas usage parameter semantic feature vector, the system can achieve more advanced functions, such as anomaly detection and early warning, so as to detect potential abnormal conditions in the gas supply pipeline, such as abnormal energy consumption or early signs of equipment failure. This early warning system helps to respond to problems in a timely manner and take necessary measures to ensure the stability and safety of gas supply.

[0026] Figure 2 FIG. 1 is a block diagram of a gas operation management data extraction module in an intelligent gas operation management system based on Ai technology according to an embodiment of the present application. Figure 2As shown, in a specific embodiment of the present application, the gas operation management data extraction module 120 includes: a gas pressure value feature extraction unit 121, which is used to perform feature extraction on the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor to obtain a gas pressure feature vector; a gas flow value feature extraction unit 122, which is used to perform feature extraction on the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor to obtain a gas flow feature vector; a gas supply pipe gas parameter feature aggregation unit 123, which is used to perform feature aggregation on the gas pressure feature vector and the gas flow feature vector to obtain a global correlation feature vector of the gas supply pipe gas parameters; a dynamic gas usage parameter feature encoding unit 124, which is used to perform feature encoding on the dynamic gas usage parameter data collected by the gas equipment parameter collector to obtain the dynamic gas usage parameter semantic feature vector.

[0027] It should be understood that the pressure data of the gas pipeline reflects the state and operation of the gas inside the pipeline. By collecting pressure values ​​at multiple predetermined time points, time series data can be constructed to capture the dynamic changes and periodic trends of the pressure inside the pipeline. These data are the basis for analyzing whether the pipeline is operating normally and whether there are any abnormal conditions. Through feature extraction, statistical features, frequency domain features or other advanced features such as peak value, mean, variance, spectral density, etc. can be extracted from these time series. These features can effectively summarize and describe the main characteristics of the pressure data. The generation of gas pressure feature vectors helps the system monitor and predict the operating status of the pipeline. This real-time monitoring and early warning capability is of great significance for ensuring the safety of pipeline operation and responding to problems in a timely manner, and helps to avoid potential safety risks and gas supply interruptions.

[0028] Furthermore, feature extraction of the gas flow values ​​collected by the gas supply pipe flow sensor is a crucial part of the intelligent gas management system. This process not only helps to monitor and optimize pipeline operation in real time, but also supports the prediction and management of the efficiency and safety of gas supply. The gas flow data reflects the actual delivery of gas inside the pipeline. Through the flow values ​​collected by the gas supply pipe flow sensor at multiple time points, the system can construct time series data, which contains the dynamic changes and trends of the gas delivery volume in the pipeline.

[0029] Furthermore, the gas pressure feature vector and the gas flow feature vector capture the dynamic changes and characteristics of pipeline pressure and flow over time, respectively, and reflect the state and characteristics of gas transportation inside the pipeline, such as pressure fluctuations, flow peaks and valleys, and load change patterns. Through feature aggregation, the gas pressure feature vector and the gas flow feature vector are combined to obtain the global correlation feature vector of gas parameters in the gas supply pipe. This global feature vector can more comprehensively describe the overall operating state and characteristics of the pipeline, not only considering the individual characteristics of pressure and flow, but also reflecting the correlation and mutual influence between them.

[0030] In particular, dynamic gas usage parameter data contains information related to user gas usage behavior, equipment working status and energy consumption. These data are obtained by real-time monitoring and recording of users' gas usage patterns, equipment operating efficiency, energy consumption, etc. Through feature coding, representative features can be extracted from these data, such as gas usage intensity, frequency, duration, changes in gas usage patterns, etc. These features can comprehensively describe users' gas usage behavior and equipment operating characteristics.

[0031] In a specific embodiment of the present application, the gas pressure value feature extraction unit 121 includes: arranging the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor according to the time dimension to obtain a gas pressure input vector; passing the gas pressure input vector through a gas supply pipe gas pressure timing encoder to obtain the gas pressure feature vector.

[0032] It should be understood that the construction of the gas pressure input vector is based on the integration and organization of time series data. By arranging the pressure data according to the time dimension, the pressure values ​​at different time points can be arranged in an orderly manner into a vector form. This arrangement retains the time relationship of the pressure data and can clearly reflect the dynamic changes and trends of the gas pressure in the pipeline. For example, through the data points collected every minute or every second, a pressure vector containing a time sequence can be formed, thereby effectively capturing the real-time changes in the pressure inside the pipeline.

[0033] Furthermore, the gas pressure input vector is a time series vector composed of gas pressure data collected at multiple time points. These data reflect the real-time changes and fluctuations of the gas pressure inside the pipeline. Through the gas supply pipe gas pressure time series encoder, the system can effectively encode and extract these time series data to generate a representative pressure feature vector. The gas pressure feature vector can include a variety of key features related to pressure through the processing of the time series encoder. These features may include the average value, maximum value, minimum value, rate of change, periodicity, etc. of the pressure. The role of the encoder is to convert the time series data into a higher-level and more abstract feature representation, which can more comprehensively and effectively describe the dynamic characteristics of the pipeline pressure. By processing the gas pressure input vector through the gas supply pipe gas pressure time series encoder to obtain the gas pressure feature vector, the system not only enhances the ability to deeply understand the pipeline pressure data, but also improves the accuracy and efficiency of pipeline operation monitoring and management decisions. Specifically, the gas pressure input vector is fully-connectedly encoded using the fully-connected layer of the gas supply pipe gas pressure timing encoder to extract high-dimensional implicit features of the eigenvalues ​​at each position in the gas pressure input vector; and the gas pressure input vector is one-dimensionally encoded using the one-dimensional convolutional layer of the gas supply pipe gas pressure timing encoder to extract high-dimensional implicit correlation features of the correlation between the eigenvalues ​​at each position in the gas pressure input vector.

[0034] In a specific embodiment of the present application, the gas flow value feature extraction unit 122 includes: constructing the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor into a gas flow input vector; and passing the gas flow input vector through a gas flow multi-scale neighborhood feature extractor to obtain the gas flow feature vector.

[0035] It should be understood that the gas flow input vector is a vector composed of gas flow data collected at continuous time points. These data record the flow rate and amount of gas in the pipeline, reflecting the intensity and fluctuation of gas consumption in different time periods. By arranging these data in a vector form in an orderly manner according to the time dimension, the system can clearly present the changing trend and periodicity of the gas flow in the pipeline. The construction of the gas flow input vector helps to effectively manage and analyze the flow data.

[0036] Furthermore, the gas flow input vector is processed by the gas flow multi-scale neighborhood feature extractor to obtain the gas flow feature vector, in order to deeply analyze the characteristics and changes of the gas flow in the pipeline from multiple levels and angles. The gas flow input vector contains the flow data collected at continuous time points, reflecting the real-time flow of gas in the pipeline. Through the gas flow multi-scale neighborhood feature extractor, the system can extract local features of the flow data at different time scales. These time scales can include windows from minutes to hours or even longer time spans to capture the change patterns and trends of the flow at different time scales. The generation of the gas flow feature vector involves the combination and synthesis of multi-scale neighborhood features. By extracting the statistical features of the flow data at each time scale, such as the mean, standard deviation, peak value, etc., the system can obtain the local feature vector at each scale. The combination of these local feature vectors constitutes the gas flow feature vector, which can more comprehensively describe the dynamic changes and trends of the gas flow in the pipeline. Specifically, the multi-scale neighborhood feature extractor for gas flow includes: a first convolution layer, a second convolution layer parallel to the first convolution layer, and a cascade layer connected to the first convolution layer and the second convolution layer, wherein the first convolution layer uses a one-dimensional convolution kernel with a first scale, and the second convolution layer uses a one-dimensional convolution kernel with a second scale. More specifically, the first convolution layer of the multi-scale neighborhood feature extractor for gas flow is used to perform one-dimensional convolution encoding on the gas flow input vector to obtain a first-scale feature vector; the second convolution layer of the multi-scale neighborhood feature extractor for gas flow is used to perform one-dimensional convolution encoding on the gas flow input vector to obtain a second-scale feature vector; the first-scale feature vector and the second-scale feature vector are cascaded to obtain the gas flow feature vector.

[0037] In a specific embodiment of the present application, the gas supply pipe gas parameter feature aggregation unit 123 includes: associating the gas pressure feature vector and the gas flow feature vector to obtain a gas supply pipe gas parameter association feature matrix; passing the gas supply pipe gas parameter association feature matrix through a gas supply pipe gas parameter feature encoder based on a convolutional neural network to obtain the gas supply pipe gas parameter global association feature vector.

[0038] It should be understood that the gas pressure eigenvector and the gas flow eigenvector reflect the dynamic characteristics of the pressure and flow inside the pipeline respectively. By correlating these two types of eigenvectors, the spatiotemporal relationship and mutual influence between pressure and flow can be revealed. The obtained gas supply pipe gas parameter correlation feature matrix provides a more comprehensive understanding of the pipeline operation status by integrating the multi-dimensional characteristics of pressure and flow. Each element in the matrix can be regarded as an indicator of the correlation between pressure and flow at a specific time and location. This correlation analysis not only helps to identify abnormal situations or potential problems, but also helps to optimize the gas supply scheduling and operation strategy of the pipeline.

[0039] Furthermore, the gas supply pipe gas parameter correlation feature matrix is ​​processed by a gas supply pipe gas parameter feature encoder based on a convolutional neural network, aiming to extract high-level, global feature vectors from complex correlation data, so as to further enhance the ability to understand and analyze the state and operating characteristics of the gas supply system. Among them, the gas supply pipe gas parameter correlation feature matrix contains correlation data between multi-dimensional parameters such as pressure and flow. These data reflect the complex interactions and dynamic changes inside the pipeline, such as how pressure changes affect flow, or the propagation pattern of pressure and flow between different positions. Through convolutional neural networks, especially convolutional structures suitable for sequence data, these spatial and temporal correlation features can be effectively captured. The gas supply pipe gas parameter feature encoder based on convolutional neural networks can extract global and abstract feature representations from the correlation feature matrix. Among them, the convolutional neural network can effectively identify and extract important features of different positions and different relationships in the matrix through the structure of convolutional layers and pooling layers, such as pressure and flow fluctuation patterns with significant influence, or the difference between abnormal modes and normal working modes. Specifically, each layer of the gas supply pipe gas parameter feature encoder based on the convolutional neural network performs convolution processing, mean pooling processing based on the local feature matrix and nonlinear activation processing on the input data in the forward transfer of the layer, so that the last layer of the gas supply pipe gas parameter feature encoder based on the convolutional neural network outputs the gas supply pipe gas parameter global correlation feature vector, wherein the input of the gas supply pipe gas parameter feature encoder based on the convolutional neural network is the gas supply pipe gas parameter correlation feature matrix.

[0040] In a specific embodiment of the present application, the dynamic gas usage parameter feature encoding unit 124 includes: segmenting the dynamic gas usage parameter data collected by the gas usage equipment parameter collector to obtain a dynamic gas usage parameter word sequence; passing the dynamic gas usage parameter word sequence through a dynamic gas usage parameter context encoder based on a convolutional neural network to obtain the dynamic gas usage parameter semantic feature vector.

[0041] It should be understood that dynamic gas consumption parameter data usually includes real-time data collected from multiple gas-using devices and sensors, such as pressure, flow, temperature, etc. These data record the operating status and environmental conditions of gas-using devices inside the pipeline in a high-frequency and multi-dimensional manner. By segmenting these data, continuous numerical data can be divided into discrete semantic units according to timestamps, such as data fragments per minute or hour. The dynamic gas consumption parameter word sequence after segmentation supports data mining and pattern recognition. By applying sequence analysis methods in natural language processing technology, such as sequence pattern mining, time series clustering or sequence prediction, hidden laws and trends in the data can be discovered. For example, identifying changes in gas consumption patterns in different seasons, synchronous operation modes between specific equipment, or prediction and detection of abnormal events.

[0042] Furthermore, the dynamic gas consumption parameter word sequence is processed by a dynamic gas consumption parameter context encoder based on a convolutional neural network, aiming to convert time series data into high-level feature representations with semantic information, thereby deepening the understanding and analysis capabilities of gas consumption behavior and system status. Among them, the dynamic gas consumption parameter word sequence reflects the changes and activities of gas consumption equipment inside the pipeline on the time axis. These word sequences are composed of semantic units segmented by timestamps, such as hourly, daily or weekly gas consumption data fragments. Through convolutional neural networks, especially convolutional neural network architectures suitable for sequence data, contextual information and time dependencies in word sequences can be effectively captured. The dynamic gas consumption parameter context encoder based on convolutional neural networks can extract semantic feature vectors of dynamic gas consumption parameters. These feature vectors not only contain the specific values ​​in each time period, but also integrate the contextual information and the changing trend of the time series. Specifically, the embedding layer of the dynamic gas usage parameter context encoder based on convolutional neural network is used to map each word in the dynamic gas usage parameter word sequence into a word embedding vector to obtain a sequence of word embedding vectors; the converter-based Bert model of the dynamic gas usage parameter context encoder based on convolutional neural network is used to perform global context semantic encoding on the sequence of word embedding vectors to obtain multiple feature vectors; and the multiple feature vectors are cascaded to obtain the dynamic gas usage parameter semantic feature vector.

[0043] In the above-mentioned intelligent gas operation and management system 100 based on Ai technology, the gas energy consumption abnormal warning judgment module 130 is used to determine whether to issue a gas energy consumption abnormal warning based on the global correlation feature vector of the gas supply pipe gas parameters and the dynamic gas use parameter semantic feature vector. It should be understood that the global correlation feature vector of the gas supply pipe gas parameters reflects the key operating characteristics of the entire pipeline system, such as the complex relationship between pressure and flow, the synergy between different positions, and the long-term stable operating mode. These feature vectors can capture the basic operating status of the pipeline system by comprehensively analyzing multidimensional data, providing important benchmarks and background information for anomaly detection. The dynamic gas use parameter semantic feature vector provides a deep understanding of gas use behavior in a specific time period. The global correlation feature vector of the gas supply pipe gas parameters and the dynamic gas use parameter semantic feature vector not only contain the numerical value of the time series data, but also integrate contextual information and semantic features, such as changes in gas use patterns, the occurrence of specific events, and changes in the operating status of equipment. By comprehensively considering the global correlation features and dynamic semantic features, the system can make a comprehensive judgment and predict the abnormal energy consumption trend that may occur in the future. This predictive analysis helps take proactive action to prevent potential system failures or reduced energy efficiency.

[0044] Figure 3 FIG. 1 is a block diagram of a gas energy consumption abnormality warning judgment module in an intelligent gas operation management system based on Ai technology according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the gas energy consumption abnormal warning judgment module 130 includes: a gas energy consumption feature fusion unit 131, which is used to fuse the global correlation feature vector of the gas parameter of the gas supply pipe and the semantic feature vector of the dynamic gas usage parameter to obtain a gas energy consumption abnormal judgment classification feature vector; a gas energy consumption feature optimization unit 132, which is used to perform a learning predictive matching optimization based on the viewpoint metric prediction contribution on the gas energy consumption abnormal judgment classification feature vector to obtain an optimized gas energy consumption abnormal judgment classification feature vector; an abnormal warning judgment classification unit 133, which is used to pass the optimized gas energy consumption abnormal judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a gas energy consumption abnormality warning.

[0045] It should be understood that by integrating the global correlation feature vector of gas supply pipe gas parameters with the semantic feature vector of dynamic gas usage parameters, the system can comprehensively analyze the long-term operation trend of the pipeline system and the dynamic changes of short-term gas usage behavior. This comprehensive analysis capability helps to identify complex abnormal patterns and unusual operating conditions that are difficult to capture with a single feature vector.

[0046] In particular, in the technical solution of the present application, it is considered that if the data collection of the gas supply pipe pressure sensor, flow sensor and gas equipment parameter collector is incomplete or there are errors, the input vector may be inaccurate, thereby affecting the quality of the feature vector and the judgment of the classifier. Feature extraction methods such as time series encoders, multi-scale neighborhood feature extractors, and context encoders may not be able to fully capture all features related to abnormal gas energy consumption, or may be too sensitive to certain features and ignore other important features. When fusing the gas pressure feature vector, the gas flow feature vector and the dynamic gas consumption parameter semantic feature vector, if the weight distribution is improper or the fusion method is inappropriate, it may lead to the loss or deviation of important information, affecting the final gas energy consumption abnormality judgment classification feature vector. When the classification judgment is made by the classifier, there will be a class harmony offset relative to the predetermined class label of the classifier, affecting the classification iteration effect. In order to reduce the class harmony offset, in the technical solution of the present application, the gas energy consumption abnormality judgment classification feature vector is optimized based on the learning predictive matching optimization of the viewpoint metric prediction contribution to obtain an optimized gas energy consumption abnormality judgment classification feature vector.

[0047] Among them, the gas energy consumption abnormality judgment classification feature vector is subjected to learning predictive matching optimization based on viewpoint metric prediction contribution to obtain an optimized gas energy consumption abnormality judgment classification feature vector, including: probabilistic activation of the gas energy consumption abnormality judgment classification feature vector based on Sigmoid function to obtain a probabilistic gas energy consumption abnormality judgment classification feature vector; inputting the gas energy consumption abnormality judgment classification feature vector into a classifier to obtain an initial class label probability value; calculating the difference between the feature value of each position in the probabilistic gas energy consumption abnormality judgment classification feature vector and the initial class label probability value to obtain a difference feature vector, and passing the difference feature vector through a mask function to obtain a viewpoint metric prediction initial weight vector; calculating the viewpoint The autocorrelation matrix of the initial weight vector of the measurement prediction is measured, and the autocorrelation matrix of the initial weight vector of the viewpoint measurement prediction is divided by the square root of the length of the initial weight vector of the viewpoint measurement prediction, and then multiplied by the initial weight vector of the viewpoint measurement prediction through the softmax function to obtain the viewpoint measurement prediction attention-oriented weight vector; the viewpoint measurement prediction attention-oriented weight vector and the gas energy consumption abnormality judgment classification feature vector are multiplied by position points to obtain a first weighted gas energy consumption abnormality judgment classification feature vector; a predetermined weight hyperparameter is multiplied by the gas energy consumption abnormality judgment classification feature vector, and then the predetermined weight hyperparameter is added to the first weighted gas energy consumption abnormality judgment classification feature vector by position to obtain an optimized gas energy consumption abnormality judgment classification feature vector.

[0048] Here, the gas energy consumption abnormality judgment classification feature vector is subjected to a learning predictive matching optimization based on the viewpoint metric prediction contribution to obtain an optimized gas energy consumption abnormality judgment classification feature vector, including:

[0049] ;

[0050] in, represents the classification feature vector for judging abnormal gas energy consumption, represents the activation function, represents the probabilistic classification feature vector for abnormal gas energy consumption judgment, represents the initial class label probability value, It means subtracting by position. represents the mask function, A variable representing the input mask function, represents a predetermined threshold, represents the initial weight vector for viewpoint metric prediction, represents matrix multiplication, represents the transpose of a vector, represents the length of the initial weight vector for viewpoint metric prediction, represents the normalized exponential function, represents the viewpoint metric prediction attention-oriented weight vector, represents the predetermined weight hyperparameter, It means point multiplication by position. It means adding by position. Represents the optimized classification feature vector for abnormal gas energy consumption judgment.

[0051] That is, in the technical solution of the present application, the gas energy consumption abnormality judgment classification feature vector is subjected to a learning predictive matching optimization based on the viewpoint metric prediction contribution. First, the gas energy consumption abnormality judgment classification feature vector is subjected to a probabilistic activation based on the Sigmoid function to obtain a probabilistic gas energy consumption abnormality judgment classification feature vector, and the gas energy consumption abnormality judgment classification feature vector is input into the classifier to obtain an initial class label probability value, and the viewpoint metric prediction contribution of each position in the standardized probabilistic gas energy consumption abnormality judgment classification feature vector to the predetermined class probability label is initialized through inverse error analysis and masking.

[0052] Then, the viewpoint space of the reverse optimization is expanded by using the autocorrelation attention representation of the initial weight vector predicted by the viewpoint metric, and its sub-dimensional consistency is used as the attention-guided weight. The reverse consistency attention mechanism of the sub-dimensional distribution is applied to the classification feature vector for judging the abnormal gas energy consumption. In this way, the learning predictive matching of the forward feature manifold and the backward class probability distribution is performed based on the viewpoint metric prediction of post-machine learning, so that the optimized classification feature vector for judging the abnormal gas energy consumption has a relatively good distribution consistency relative to the class probability distribution of the classifier, so as to improve the accuracy of its classification judgment through the classifier.

[0053] Furthermore, the optimized gas energy consumption abnormality judgment classification feature vector is processed by a classifier to obtain a classification result, and further used to determine whether to issue a gas energy consumption abnormality warning. This is to achieve automatic monitoring and early warning of the complex gas system status through machine learning and intelligent analysis methods, thereby improving the efficiency and reliability of energy management. Among them, the classifier can quickly process the data generated in real time to achieve real-time monitoring and analysis of the gas energy consumption status. This immediacy can help the system to promptly discover and respond to potential problems, avoiding economic losses or equipment damage caused by abnormal energy consumption. The application of this technology not only improves the monitoring and management level of the energy system, but also provides strong support for the sustainable development and resource utilization optimization of smart cities.

[0054] In summary, the embodiment of the present application first obtains the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas consumption parameter data collected by the gas equipment parameter collector, and then uses deep learning technology to perform feature extraction and correlation analysis on the three. Finally, the classification result is obtained through the classifier to determine whether to issue an abnormal gas energy consumption warning, thereby optimizing energy efficiency, improving the effective utilization rate of energy, effectively controlling costs, and avoiding unnecessary expenses caused by energy waste.

[0055] As described above, the intelligent gas operation and management system 100 based on Ai technology according to the embodiment of the present application can be implemented in various terminal devices. In one example, the intelligent gas operation and management system 100 based on Ai technology can be integrated into the terminal device as a software module and / or a hardware module. For example, the intelligent gas operation and management system 100 based on Ai technology can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the intelligent gas operation and management system 100 based on Ai technology can also be one of the many hardware modules of the terminal device.

[0056] Alternatively, in another example, the intelligent gas operation and management system 100 based on Ai technology and the terminal device may also be separate devices, and the intelligent gas operation and management system 100 based on Ai technology may be connected to the terminal device via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0057] Figure 4 Flow chart of the intelligent gas operation management method based on AI technology according to an embodiment of the present application. Figure 4 As shown, according to the intelligent gas operation and management method based on Ai technology according to the embodiment of the present application, it includes: S110, obtaining the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas equipment parameter collector; S120, extracting the global correlation feature vector of the gas supply pipe gas parameters and the dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas equipment parameter collector; S130, judging whether to issue an abnormal gas energy consumption warning based on the global correlation feature vector of the gas supply pipe gas parameters and the dynamic gas usage parameter semantic feature vector.

[0058] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned intelligent gas operation management method based on Ai technology have been referred to above. Figures 1 to 3 The description of the intelligent gas operation and management system based on Ai technology has been introduced in detail, and therefore, its repeated description will be omitted.

[0059] Below, reference Figure 5 To describe an electronic device according to an embodiment of the present application.

[0060] like Figure 5 As shown, the electronic device 10 includes an input device 11, an input interface 12, a central processing unit 13, a memory 14, an output interface 15, an output device 16, and a bus 17. The input interface 12, the central processing unit 13, the memory 14, and the output interface 15 are connected to each other through the bus 17, and the input device 11 and the output device 16 are connected to the bus 17 through the input interface 12 and the output interface 15, respectively, and then connected to other components of the electronic device 10.

[0061] Specifically, the input device 11 receives input information from the outside and transmits the input information to the central processing unit 13 through the input interface 12; the central processing unit 13 processes the input information based on the computer executable instructions stored in the memory 14 to generate output information, temporarily or permanently stores the output information in the memory 14, and then transmits the output information to the output device 16 through the output interface 15; the output device 16 outputs the output information to the outside of the electronic device 10 for user use.

[0062] In one embodiment, Figure 5 The electronic device 10 shown can be implemented as a network device, which may include: a memory configured to store programs; a processor configured to run the programs stored in the memory to execute any one of the smart gas operation and management methods based on Ai technology described in the above embodiments.

[0063] According to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network, and / or installed from a removable storage medium.

[0064] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0065] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present application, but the present application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of the present application, and these modifications and improvements are also considered to be within the scope of protection of the present application.

Claims

1. An intelligent gas operation and management system based on Ai technology, characterized in that: include: The gas operation management data acquisition module is used to acquire the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector; A gas operation management data extraction module, used to extract a global correlation feature vector of gas supply pipe gas parameters and a dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector; A gas energy consumption abnormality warning judgment module, used to judge whether to issue a gas energy consumption abnormality warning based on the gas supply pipe gas parameter global correlation feature vector and the dynamic gas usage parameter semantic feature vector; The gas energy consumption abnormality warning and judgment module includes: A gas energy consumption feature fusion unit, used for fusing the gas supply pipe gas parameter global correlation feature vector and the dynamic gas usage parameter semantic feature vector to obtain a gas energy consumption abnormality judgment classification feature vector; A gas energy consumption feature optimization unit, used for performing a learning predictive matching optimization based on a viewpoint metric prediction contribution on the gas energy consumption abnormality judgment classification feature vector to obtain an optimized gas energy consumption abnormality judgment classification feature vector; An abnormal warning judgment classification unit is used to pass the optimized gas energy consumption abnormal judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a gas energy consumption abnormality warning; Wherein, the gas energy consumption characteristic optimization unit comprises: Performing probabilistic activation on the gas energy consumption abnormality judgment classification feature vector based on the Sigmoid function to obtain a probabilistic gas energy consumption abnormality judgment classification feature vector; Inputting the gas energy consumption abnormality judgment classification feature vector into a classifier to obtain an initial class label probability value; Calculate the difference between the feature value of each position in the probabilistic gas energy consumption abnormality judgment classification feature vector and the initial class label probability value to obtain a difference feature vector, and pass the difference feature vector through a mask function to obtain an initial weight vector for viewpoint metric prediction; Calculating the autocorrelation matrix of the viewpoint metric prediction initial weight vector, dividing the autocorrelation matrix of the viewpoint metric prediction initial weight vector by the square root of the length of the viewpoint metric prediction initial weight vector, and then multiplying it by the viewpoint metric prediction initial weight vector through a softmax function to obtain a viewpoint metric prediction attention-oriented weight vector; Multiplying the viewpoint metric prediction attention orientation weight vector by the gas energy consumption abnormality judgment classification feature vector by position point to obtain a first weighted gas energy consumption abnormality judgment classification feature vector; The predetermined weight hyperparameter is multiplied by the gas energy consumption abnormality judgment classification feature vector, and then the multiplication is performed with the first weighted gas energy consumption abnormality judgment classification feature vector by position to obtain an optimized gas energy consumption abnormality judgment classification feature vector.

2. The intelligent gas operation and management system based on Ai technology according to claim 1 is characterized in that: The gas operation management data extraction module includes: A gas pressure value feature extraction unit, used for extracting features from the gas pressure values ​​at a plurality of predetermined time points collected by the gas supply pipe pressure sensor to obtain a gas pressure feature vector; A gas flow value feature extraction unit, used for extracting features from the gas flow values ​​at a plurality of predetermined time points collected by the gas supply pipe flow sensor to obtain a gas flow feature vector; A gas supply pipe gas parameter feature aggregation unit, used for performing feature aggregation on the gas pressure feature vector and the gas flow feature vector to obtain a global correlation feature vector of the gas supply pipe gas parameter; The dynamic gas usage parameter feature encoding unit is used to perform feature encoding on the dynamic gas usage parameter data collected by the gas usage equipment parameter collector to obtain the dynamic gas usage parameter semantic feature vector.

3. The intelligent gas operation and management system based on Ai technology according to claim 2 is characterized in that: The gas pressure value feature extraction unit comprises: Arranging the gas pressure values ​​at a plurality of predetermined time points collected by the gas supply pipe pressure sensor according to the time dimension to obtain a gas pressure input vector; The gas pressure input vector is passed through the gas supply pipe gas pressure timing encoder to obtain the gas pressure characteristic vector.

4. The intelligent gas operation and management system based on Ai technology according to claim 3 is characterized in that: The gas flow value feature extraction unit includes: Constructing the gas flow values ​​at a plurality of predetermined time points collected by the gas supply pipe flow sensor into a gas flow input vector; The gas flow input vector is passed through a gas flow multi-scale neighborhood feature extractor to obtain the gas flow feature vector.

5. The intelligent gas operation and management system based on Ai technology according to claim 4 is characterized in that: The gas supply pipe gas parameter feature aggregation unit comprises: Correlating the gas pressure characteristic vector with the gas flow characteristic vector to obtain a gas supply pipe gas parameter correlation characteristic matrix; The gas supply pipe gas parameter associated feature matrix is ​​passed through a gas supply pipe gas parameter feature encoder based on a convolutional neural network to obtain the gas supply pipe gas parameter global associated feature vector.

6. The intelligent gas operation and management system based on Ai technology according to claim 5 is characterized in that: The dynamic gas usage parameter characteristic encoding unit comprises: Segmenting the dynamic gas usage parameter data collected by the gas usage equipment parameter collector to obtain a dynamic gas usage parameter word sequence; The dynamic gas usage parameter word sequence is passed through a dynamic gas usage parameter context encoder based on a convolutional neural network to obtain the dynamic gas usage parameter semantic feature vector.

7. An intelligent gas operation management method based on Ai technology, characterized in that: include: Obtaining gas pressure values ​​at multiple predetermined time points collected by a gas supply pipe pressure sensor, gas flow values ​​at multiple predetermined time points collected by a gas supply pipe flow sensor, and dynamic gas usage parameter data collected by a gas usage equipment parameter collector; Extracting a gas supply pipe gas parameter global correlation feature vector and a dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector; Based on the global correlation feature vector of the gas supply pipe gas parameter and the semantic feature vector of the dynamic gas usage parameter, determining whether to issue an abnormal gas energy consumption warning; Wherein, judging whether to issue an abnormal gas energy consumption warning based on the global correlation feature vector of the gas supply pipe gas parameter and the semantic feature vector of the dynamic gas usage parameter includes: The global correlation feature vector of the gas supply pipe gas parameter and the semantic feature vector of the dynamic gas consumption parameter are merged to obtain a classification feature vector for judging abnormal gas energy consumption; The gas energy consumption abnormality judgment classification feature vector is subjected to a learning predictive matching optimization based on a viewpoint metric prediction contribution to obtain an optimized gas energy consumption abnormality judgment classification feature vector; The optimized gas energy consumption abnormality judgment classification feature vector is passed through a classifier to obtain a classification result, and the classification result is used to determine whether to issue a gas energy consumption abnormality warning; The gas energy consumption abnormality judgment classification feature vector is subjected to a learning predictive matching optimization based on a viewpoint metric prediction contribution to obtain an optimized gas energy consumption abnormality judgment classification feature vector, including: Performing probabilistic activation on the gas energy consumption abnormality judgment classification feature vector based on the Sigmoid function to obtain a probabilistic gas energy consumption abnormality judgment classification feature vector; Inputting the gas energy consumption abnormality judgment classification feature vector into a classifier to obtain an initial class label probability value; Calculate the difference between the feature value of each position in the probabilistic gas energy consumption abnormality judgment classification feature vector and the initial class label probability value to obtain a difference feature vector, and pass the difference feature vector through a mask function to obtain an initial weight vector for viewpoint metric prediction; Calculating the autocorrelation matrix of the viewpoint metric prediction initial weight vector, dividing the autocorrelation matrix of the viewpoint metric prediction initial weight vector by the square root of the length of the viewpoint metric prediction initial weight vector, and then multiplying it by the viewpoint metric prediction initial weight vector through a softmax function to obtain a viewpoint metric prediction attention-oriented weight vector; Multiplying the viewpoint metric prediction attention orientation weight vector by the gas energy consumption abnormality judgment classification feature vector by position point to obtain a first weighted gas energy consumption abnormality judgment classification feature vector; The predetermined weight hyperparameter is multiplied by the gas energy consumption abnormality judgment classification feature vector, and then the multiplication is performed with the first weighted gas energy consumption abnormality judgment classification feature vector by position to obtain an optimized gas energy consumption abnormality judgment classification feature vector.

8. The intelligent gas operation management method based on Ai technology according to claim 7 is characterized in that: Extracting a gas supply pipe gas parameter global association feature vector and a dynamic gas usage parameter semantic feature vector from the gas pressure values ​​at multiple predetermined time points collected by the gas supply pipe pressure sensor, the gas flow values ​​at multiple predetermined time points collected by the gas supply pipe flow sensor, and the dynamic gas usage parameter data collected by the gas usage equipment parameter collector, including: Extracting features of the gas pressure values ​​at a plurality of predetermined time points collected by the gas supply pipe pressure sensor to obtain a gas pressure feature vector; Extracting features of the gas flow values ​​at a plurality of predetermined time points collected by the gas supply pipe flow sensor to obtain a gas flow feature vector; Performing feature aggregation on the gas pressure feature vector and the gas flow feature vector to obtain a global correlation feature vector of the gas supply pipe gas parameters; The dynamic gas usage parameter data collected by the gas usage equipment parameter collector is feature encoded to obtain the dynamic gas usage parameter semantic feature vector.

Images