Abnormal event early warning method and system for monitoring gas leakage

Abstract

The invention belongs to the technical field of abnormal early warning, and discloses an abnormal event early warning method and system for monitoring gas leakage. The method comprises the following steps: constructing a gas leakage monitoring digital twinborn model and an abnormal event early warning model based on a cloud data center; based on the gas leakage monitoring equipment, collecting real-time gas leakage monitoring data of the monitoring point at each collection moment, and uploading the real-time gas leakage monitoring data to the cloud data center; performing abnormal event early warning by using an abnormal event early warning model based on the cloud data center according to the plurality of real-time gas leakage monitoring data at the same acquisition moment to obtain a real-time abnormal event early warning result; and according to a real-time abnormal event early warning result, performing abnormal event early warning visualization by using a gas leakage monitoring digital twinborn model, and controlling corresponding gas leakage response equipment. According to the invention, the problems of low intelligent degree, poor accuracy, low practicability, poor early warning effect and lack of response mechanism in the prior art are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of abnormality early warning, and in particular relates to an abnormal event early warning method and system for monitoring gas leakage. Background Art

[0002] Abnormal event warnings analyze and judge monitoring data to identify potential abnormal situations and issue timely warnings. This helps to control the development of abnormal situations in the early stages of safety accidents or during the incubation period of safety hazards, protecting people, property, and production from damage. Especially in gas leak monitoring tasks, reliable and accurate abnormal event warnings can avoid serious safety accidents caused by gas leaks. Existing abnormal event warning technologies for monitoring gas leaks have the following shortcomings:

[0003] 1) Monitoring is performed using simple trigger-type gas sensors, which have low intelligence and accuracy, and are unable to provide timely warnings for abnormal events caused by gas leaks, resulting in low practicality;

[0004] 2) The warning effect of abnormal events is poor. It only uses alarm signals to issue safety warnings, but cannot accurately find the location of gas leaks. In addition, there is a lack of response mechanism, and it is impossible to provide accurate and appropriate response strategies for the abnormal events warned. Summary of the Invention

[0005] In order to solve the problems of low intelligence, poor accuracy, low practicality, poor warning effect and lack of response mechanism in the existing technology, the purpose of the present invention is to provide a method and system for monitoring abnormal event warning of gas leakage.

[0006] The technical solution adopted in the present invention is:

[0007] A method for monitoring abnormal gas leakage and early warning includes the following steps:

[0008] Based on the cloud data center, digital twin technology is used to build a digital twin model for gas leak monitoring, and artificial intelligence algorithms are used to build an abnormal event early warning model;

[0009] Based on the gas leakage monitoring equipment, real-time gas leakage monitoring data of the monitoring point is collected at each collection moment, and the real-time gas leakage monitoring data is uploaded to the cloud data center;

[0010] Based on the cloud data center, the abnormal event warning model is used to issue abnormal event warnings based on multiple real-time gas leakage monitoring data collected at the same time, and obtain real-time abnormal event warning results;

[0011] Based on the cloud data center, according to the real-time abnormal event warning results, the gas leakage monitoring digital twin model is used to visualize the abnormal event warning and control the corresponding gas leakage response equipment.

[0012] Furthermore, the abnormal event warning model includes a gas leakage analysis model, a gas leakage location model, and an abnormal event response model;

[0013] Real-time abnormal event warning results include real-time gas leakage analysis results, real-time gas leakage location results, and real-time abnormal event response strategies;

[0014] The real-time abnormal event response strategy includes a real-time abnormal event warning signal and a real-time gas leakage response instruction of the gas leakage response equipment.

[0015] Furthermore, based on the cloud data center, digital twin technology is used to build a gas leak monitoring digital twin model, and artificial intelligence algorithms are used to build an abnormal event warning model, including the following steps:

[0016] Based on the topology information of the gas transmission pipeline network, a static sub-model of the gas transmission pipeline network topology is constructed using digital twin technology based on the cloud data center.

[0017] Based on the first basic attribute information of all gas leakage monitoring equipment, using digital twin technology, set the corresponding gas leakage monitoring equipment dynamic sub-model at the corresponding position of the gas transmission pipeline network topology static sub-model;

[0018] Based on the second basic attribute information of all gas leakage response equipment, digital twin technology is used to set the corresponding gas leakage response equipment dynamic sub-model at the corresponding position of the gas transmission pipeline network topology static sub-model;

[0019] Integrate the static sub-model of the gas transmission pipeline network topology, the dynamic sub-models of several gas leakage monitoring devices, and the dynamic sub-models of several gas leakage response devices to obtain a gas leakage monitoring digital twin model;

[0020] Based on some historical gas leakage monitoring data, a deep learning algorithm is used to build a gas leakage analysis model and generate some historical gas leakage analysis results;

[0021] Based on the historical gas leakage analysis results, a number of historical gas leakage monitoring data and gas leakage monitoring digital twin models that indicate gas leakage were present were used to construct a gas leakage location model using a deep learning algorithm, and generate several historical gas leakage location results.

[0022] Based on several historical gas leakage analysis results and the corresponding historical gas leakage location results, a reinforcement learning algorithm is used to build an abnormal event response model and generate several historical abnormal event response experiences.

[0023] Furthermore, the gas leakage analysis model is constructed based on the RF-LSTM algorithm;

[0024] The gas leak location model is built based on the GNN-Attention-LSTM algorithm;

[0025] The abnormal event response model is built based on the DQN algorithm.

[0026] Furthermore, based on the cloud data center, based on a number of real-time gas leakage monitoring data collected at the same time, an abnormal event warning model is used to perform abnormal event warning, and a real-time abnormal event warning result is obtained, including the following steps:

[0027] Based on the cloud data center, a gas leakage analysis model is used to analyze gas leakage based on multiple real-time gas leakage monitoring data collected at the same time, and obtain real-time gas leakage analysis results.

[0028] If the real-time gas leakage analysis result indicates that there is a gas leakage, the process proceeds to the next step; otherwise, the process waits for the next collection moment of some real-time gas leakage monitoring data;

[0029] Based on a number of real-time gas leakage monitoring data and a gas leakage monitoring digital twin model, a gas leakage positioning model is used to locate gas leakage and obtain real-time gas leakage positioning results.

[0030] According to the real-time gas leakage analysis results and the corresponding real-time gas leakage positioning results, an abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy.

[0031] Furthermore, based on the cloud data center, a gas leakage analysis model is used to perform gas leakage analysis based on a plurality of real-time gas leakage monitoring data at the same collection time to obtain a real-time gas leakage analysis result, including the following steps:

[0032] Based on the cloud data center, multiple real-time gas leakage monitoring data at the same collection time are converted into a real-time gas leakage monitoring data matrix;

[0033] Using the gas leakage analysis model, M real-time key features of the real-time gas leakage monitoring data matrix are extracted, where M is the total number of key features;

[0034] Based on the M real-time key features, gas leakage analysis is performed to obtain real-time gas leakage analysis results.

[0035] Furthermore, based on a number of real-time gas leakage monitoring data and a gas leakage monitoring digital twin model, a gas leakage positioning model is used to locate the gas leakage and obtain a real-time gas leakage positioning result, including the following steps:

[0036] Using the gas leakage location model, the real-time gas transmission pipeline network topology features of the static sub-model of the gas transmission pipeline network topology in the gas leakage monitoring digital twin model are extracted;

[0037] According to the preset attention weights, the M real-time key features of the real-time gas leakage monitoring data matrix and the real-time gas transmission pipeline network topology features are spliced ​​to obtain the real-time splicing features;

[0038] Based on the real-time splicing features, gas leak positioning is performed to obtain real-time gas leak positioning results.

[0039] Furthermore, based on the real-time gas leak analysis results and the corresponding real-time gas leak location results, an abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy, which includes the following steps:

[0040] Based on the real-time gas leak analysis results and the corresponding real-time gas leak location results, a search is performed in several historical abnormal event response experiences to obtain several matching historical abnormal event response experiences;

[0041] Based on the real-time gas leakage analysis results, the real-time gas leakage location results, and a number of matching historical abnormal event response experiences, the model parameters of the abnormal event response model are updated to obtain updated model parameters;

[0042] According to the updated model parameters, the abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy.

[0043] Furthermore, based on the cloud data center and the real-time abnormal event warning results, the gas leak monitoring digital twin model is used to visualize the abnormal event warning and control the corresponding gas leak response equipment, including the following steps:

[0044] Based on the cloud data center, the real-time gas leak location results are used to locate and highlight the real-time leak point in the static sub-model of the gas transmission pipeline network topology of the gas leak monitoring digital twin model.

[0045] Add and visualize real-time gas leakage analysis results and real-time abnormal event warning signals of the real-time abnormal event response strategy to the dynamic sub-model of the gas leakage monitoring equipment closest to the real-time leakage point;

[0046] Add and visualize corresponding real-time gas leakage response instructions in the dynamic sub-models of all gas leakage response devices involved in the real-time abnormal event response strategy;

[0047] The real-time gas leakage response instruction is sent to the gas leakage response device at the corresponding response point, and the corresponding gas leakage response device is controlled according to the real-time gas leakage response instruction.

[0048] A gas leakage monitoring abnormal event early warning system is used to implement a gas leakage monitoring abnormal event early warning method. The system includes a cloud data center, a gas transmission pipeline network, a plurality of gas leakage monitoring devices and a plurality of gas leakage response devices. The cloud data center is respectively communicated with the plurality of gas leakage monitoring devices and the plurality of gas leakage response devices. The gas transmission pipeline network is provided with a plurality of monitoring points and a plurality of response points. The plurality of gas leakage monitoring devices are respectively provided at the plurality of monitoring points of the gas transmission pipeline network. The plurality of gas leakage response devices are respectively provided at the plurality of response points of the gas transmission pipeline network.

[0049] The beneficial effects of the present invention are:

[0050] The present invention discloses a method and system for monitoring abnormal event warning for gas leakage, applies digital twin technology to gas leakage monitoring, and simulates the actual operating status of the gas transmission pipeline network by constructing a digital twin model for gas leakage monitoring, thereby improving the intuitiveness and convenience of abnormal event warning. An abnormal event warning model is constructed using an artificial intelligence algorithm, including a gas leakage analysis model, a gas leakage positioning model, and an abnormal event response model, thereby realizing intelligent monitoring and early warning of gas leakage, accurately locating the location of the gas leakage, improving the abnormal event warning effect, and providing a response mechanism, providing accurate and appropriate response strategies for the abnormal events warned, thereby improving the intelligence, efficiency and accuracy of gas leakage monitoring; unified data analysis and abnormal event warning visualization are performed based on the cloud data center, thereby improving the real-time performance of gas leakage monitoring and the abnormal event warning response speed.

[0051] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flow chart of the abnormal event early warning method for monitoring gas leakage in the present invention.

[0053] Figure 2 It is a structural block diagram of the abnormal event early warning system for monitoring gas leakage in the present invention. DETAILED DESCRIPTION

[0054] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0055] Example 1:

[0056] like Figure 1 As shown, this embodiment provides a method for monitoring abnormal events and early warning of gas leakage, comprising the following steps:

[0057] S1: Based on the cloud data center, digital twin technology is used to build a gas leak monitoring digital twin model, and artificial intelligence algorithms are used to build an abnormal event warning model. The steps include:

[0058] S1-1: Based on the cloud data center and the topology information of the gas transmission pipeline network, a static sub-model of the gas transmission pipeline network topology is constructed using digital twin technology;

[0059] S1-2: Based on the first basic attribute information of all gas leakage monitoring devices, use digital twin technology to set corresponding dynamic sub-models of gas leakage monitoring devices at corresponding positions of the static sub-model of the gas transmission pipeline network topology;

[0060] S1-3: Based on the second basic attribute information of all gas leakage response equipment, use digital twin technology to set the corresponding gas leakage response equipment dynamic sub-model at the corresponding position of the gas transmission pipeline network topology static sub-model;

[0061] S1-4: Integrate the static sub-model of the gas transmission pipeline network topology, the dynamic sub-models of several gas leakage monitoring devices, and the dynamic sub-models of several gas leakage response devices to obtain a gas leakage monitoring digital twin model;

[0062] S1-5: Based on some historical gas leakage monitoring data, a gas leakage analysis model is constructed using the Random Forest (RF)-Long Short-Term Memory (LSTM) algorithm, and several historical gas leakage analysis results are generated.

[0063] The RF structure uses an internal Classification and Regression Tree (CART) to screen key features of the input data and extract several key features related to the data prediction label. The LSTM network uses a multi-layer structure to explore the potential connection between key features and data prediction labels. It trains its own network structure with a large number of key features to predict new data features, achieving fast, efficient, and accurate prediction of data labels.

[0064] S1-6: Based on the historical gas leakage analysis results, several historical gas leakage monitoring data and gas leakage monitoring digital twin models are used to build a gas leakage location model using the Graph Neural Networks (GNN)-Attention-LSTM algorithm, and generate several historical gas leakage location results.

[0065] The GNN network extracts the topological features of the gas transmission pipeline network from the static sub-model of the gas transmission pipeline network topology, including node features at the connection and turning points of the transmission pipelines, edge features between transmission pipelines, and graph features of the gas transmission pipeline network. The topological features extracted by the GNN network can accurately extract the topological information of the gas transmission pipeline network. The Attention mechanism uses preset attention weights to combine the topological features of the gas transmission pipeline network with several key features extracted by the gas leakage analysis model, integrating scattered features and enhancing the influence of features on location label prediction. The LSTM network achieves accurate location label prediction based on the combined features, reflecting the exact location of the gas leak.

[0066] S1-7: Based on several historical gas leak analysis results and corresponding historical gas leak location results, use the Deep Q Network (DQN) algorithm to build an abnormal event response model and generate several historical abnormal event response experiences, including the following steps:

[0067] S1-7-1: Use the abnormal event response strategy generation problem as a simulation environment for the reinforcement learning algorithm to build an intelligent agent and experience replay pool;

[0068] S1-7-2: Based on the gas leakage status contained in the historical gas leakage analysis results and the corresponding historical gas leakage location results, define the state space of the reinforcement learning algorithm; the state parameters in the state space include "gas leakage degree, abnormal event level, gas leakage diffusion direction, gas leakage diffusion speed, gas leakage point location, etc."

[0069] S1-7-3: Define the action space of the reinforcement learning algorithm based on the preset actions included in the abnormal event response strategy, and define the reward function based on the action space; the action parameters in the action space include "closing the gas transmission pipeline valve, responding to the abnormal event warning, turning on the ventilation equipment, activating the alarm, etc."

[0070] S1-7-4: Construct a network structure including an input layer and an output layer of a deep Q network, connect the input layer to the state space, and connect the output layer to the action space;

[0071] S1-7-5: Build an initial abnormal event response model based on the action space, state space, reward function, deep Q network, agent, and experience replay pool;

[0072] S1-7-6: Based on several historical gas leak analysis results and corresponding historical gas leak location results, optimize and train the deep Q network and intelligent agent of the initial abnormal event response model to obtain the final abnormal event response model and generate several historical abnormal event response experiences;

[0073] S1-7-7: Store several historical abnormal event response experiences into the experience replay pool of the final abnormal event response model;

[0074] S2: Based on the gas leakage monitoring equipment, real-time gas leakage monitoring data of the monitoring point is collected at each collection moment, and the real-time gas leakage monitoring data is uploaded to the cloud data center;

[0075] The gas leak monitoring equipment includes a main control module, a wireless transmission module, a gas concentration sensor, an ambient temperature sensor, an ambient wind speed sensor, an ambient humidity sensor, a pipeline vibration sensor, and a pipeline pressure sensor, all installed at the gas transmission pipeline monitoring point. These sensors collect data related to gas leaks, including gas concentration, ambient temperature, ambient wind speed, ambient humidity, pipeline vibration, and pipeline pressure. These data are crucial for determining gas leaks, their spread, and severity. The main control module controls the normal operation of other modules, while the wireless transmission module uploads the gas leak monitoring data, comprised of these indicators, to the cloud data center.

[0076] S3: Based on the cloud data center, the abnormal event warning model is used to perform abnormal event warning based on multiple real-time gas leakage monitoring data collected at the same time, and obtain real-time abnormal event warning results, including the following steps:

[0077] S3-1: Based on the cloud data center, a gas leakage analysis model is used to perform gas leakage analysis based on multiple real-time gas leakage monitoring data collected at the same time, and obtain real-time gas leakage analysis results, including the following steps:

[0078] S3-1-1: Based on the cloud data center, convert multiple real-time gas leakage monitoring data at the same collection time into a real-time gas leakage monitoring data matrix;

[0079] The column vector of the real-time gas leakage monitoring data matrix is ​​a number of real-time gas leakage monitoring data, and the row vector of the real-time gas leakage monitoring data matrix is ​​a number of real-time monitoring gas leakage indicator data of each real-time gas leakage monitoring data;

[0080] S3-1-2: Use the gas leakage analysis model and the trained RF structure to extract M real-time key features from the real-time gas leakage monitoring data matrix, where M is the total number of key features. The steps include:

[0081] S3-1-2-1: Use the RF structure trained in the key feature extraction module to extract the feature contribution of several row vector candidate features in the real-time gas leakage monitoring data matrix;

[0082] The formula is:

[0083]

[0084] Where, is the feature contribution of the candidate features of the j'th row vector; is the contribution of the j'th row vector candidate feature to the cth tree in the random forest; c is the CART tree indicator; j' is the row vector candidate feature indicator; n' is the total number of CARTs;

[0085]

[0086] Where, GI m GI l' GI r" is the Gini index of CART tree node m, node l' and node r" of random forest; p mk" is the proportion of category k" in CART tree node m; K' is the total number of categories; m, l', r" are node indicators; k" is the category indicator;

[0087] S3-1-2-2: Normalize the feature contributions of several row vector candidate features to obtain corresponding normalized feature contributions;

[0088] The formula is:

[0089]

[0090] Where VIM j' is the contribution of the feature after normalization; J is the total number of candidate features of the row vector;

[0091] S3-1-2-3: Generate the feature selection standard values ​​of several row vector candidate features based on the normalized feature contribution;

[0092] The formula is:

[0093]

[0094] Where, CFC j' The standard value for feature selection of candidate features for the j'th row vector; is the normalized feature contribution of the j*th row vector candidate feature; j* is the indicator of the row vector candidate feature;

[0095] S3-1-2-4: Sort the candidate features of the row vectors in descending order according to the feature selection standard value, and select the first M candidate features of the row vectors as real-time key features to obtain M real-time key features;

[0096] The RF structure can select M key row vector features related to gas leakage analysis and prediction from a large number of indicator data (gas concentration, ambient temperature, ambient wind speed, ambient humidity, pipeline vibration, and pipeline pressure, etc.);

[0097] S3-1-3: Based on the M real-time key features, use the first LSTM network to perform gas leakage analysis and obtain real-time gas leakage analysis results;

[0098] Real-time gas leakage analysis results include normal and gas leakage. Gas leakage includes slight gas leakage, moderate gas leakage, and severe gas leakage. Each level of gas leakage includes spreading to residential areas, spreading to industrial areas, spreading to uninhabited areas, fast gas leakage diffusion speed, slow gas leakage diffusion speed, medium gas leakage diffusion speed, and comprehensive abnormal event level.

[0099] S3-2: If the real-time gas leakage analysis result indicates that there is a gas leakage, proceed to the next step; otherwise, wait for some real-time gas leakage monitoring data at the next collection time;

[0100] S3-3: Based on some real-time gas leakage monitoring data and the gas leakage monitoring digital twin model, use the gas leakage positioning model to locate the gas leakage and obtain the real-time gas leakage positioning result, including the following steps:

[0101] S3-3-1: Use the GNN network of the gas leak location model to extract the real-time gas transmission pipeline network topology features of the static sub-model of the gas transmission pipeline network topology in the gas leak monitoring digital twin model;

[0102] S3-3-2: Based on the attention weights preset by the Attention mechanism, the M real-time key features of the real-time gas leakage monitoring data matrix and the real-time gas transmission pipeline network topology features are spliced ​​to obtain real-time splicing features;

[0103] S3-3-3: Based on the real-time splicing features, the second LSTM network is used to locate the gas leak and obtain the real-time gas leak location result;

[0104] S3-4: Based on the real-time gas leak analysis results and the corresponding real-time gas leak location results, use the abnormal event response model to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy, including the following steps:

[0105] S3-4-1: Based on the real-time gas leak analysis results and the corresponding real-time gas leak location results, search through several historical abnormal event response experiences to obtain several matching historical abnormal event response experiences;

[0106] S3-4-2: Based on the real-time gas leak analysis results, the real-time gas leak location results, and several matching historical abnormal event response experiences, the model parameters of the abnormal event response model are updated to obtain the updated model parameters, including the following steps:

[0107] S3-4-2-1: Based on the real-time gas leakage analysis results and the corresponding real-time gas leakage location results, update the state space of the abnormal event response model to obtain the updated state space S'=[s'1,...,s' i" ,...,s' I' ], where s' i" is the updated i-th state value, i" is the state indicator, and I' is the total number of state space dimensions;

[0108] S3-4-2-2: Based on some matching historical abnormal event response experiences, update the action space of the abnormal event response model to obtain the updated action space A'=[a'1,...,a' j" ,...,a' I ], where a' j" is the updated j-th "action value", j" is the action indicator, and I is the total number of action space dimensions;

[0109] S3-4-3: Based on the updated model parameters, use the abnormal event response model to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy, including the following steps:

[0110] S3-4-3-1: Input the updated state space into the deep Q network of the abnormal event response model, and use the deep Q network to generate the Q value of each possible action in the updated action space based on the updated state space;

[0111] S3-4-3-2: Use the agent to obtain the reward value of each possible action in the updated action space according to the reward function, and update the Q value of the possible action according to the reward value to obtain the updated Q value of each possible action;

[0112] The formula is:

[0113] Q(s' p',a' p' )=(1-α")·Q(s p' ,a p' )+α"·(R(s p' ,a p' ,s' p' )+γ·Q max (s p' ,a p' ))

[0114] Where, Q(s' p' ,a' p' ) is the updated state value s' p' and updated action value a' p' The corresponding updated Q value; Q(s p' ,a p' ) is the state value s p' and action value a p' The corresponding predicted Q value; α" is the learning rate; Q max (s p' ,a p' ) is the highest predicted Q value; p' is the comprehensive indicator; γ is the update parameter;

[0115] S3-4-3-3: Repeat the above Q-value update steps until the number of iterations reaches the iteration threshold, and use the greedy strategy to select the possible action with the highest updated Q-value in each iteration as the execution action;

[0116] S3-4-3-4: Use an agent to integrate several execution actions to obtain a real-time abnormal event response strategy;

[0117] The real-time abnormal event response strategy includes real-time abnormal event warning signals and real-time gas leak response instructions for gas leak response equipment. The real-time gas leak response instructions include "closing the valves of the No. 1 and No. 2 gas transmission pipelines at both ends of the leak point, responding to the abnormal event warning, and opening the No. 1 and No. 2 ventilation equipment near the gas transmission pipeline where the leak point is located";

[0118] S4: Based on the real-time abnormal event warning results, the gas leak monitoring digital twin model is used in the cloud data center to visualize abnormal event warnings and control the corresponding gas leak response equipment. This includes the following steps:

[0119] S4-1: Based on the real-time gas leak location results of the gas leak in the cloud data center, the real-time leak point is located and highlighted in the static sub-model of the gas transmission pipeline network topology of the gas leak monitoring digital twin model;

[0120] S4-2: Add and visualize the real-time gas leakage analysis results and real-time abnormal event warning signals of the real-time abnormal event response strategy to the dynamic sub-model of the gas leakage monitoring equipment closest to the real-time leakage point;

[0121] S4-3: Add and visualize corresponding real-time gas leakage response instructions in the dynamic sub-models of all gas leakage response devices involved in the real-time abnormal event response strategy;

[0122] S4-4: sending a real-time gas leakage response instruction to a gas leakage response device at a corresponding response point, and controlling the corresponding gas leakage response device according to the real-time gas leakage response instruction;

[0123] The gas leak response equipment includes alarms, remote servers, ventilation equipment, pipeline valves, safety passage signs, etc. In this embodiment, the gas leak response equipment at the response point includes gas transmission pipeline valves No. 1 and No. 2, all remote servers, ventilation equipment No. 1 and No. 2, etc.

[0124] Example 2:

[0125] like Figure 2 As shown, this embodiment provides an abnormal event warning system for monitoring gas leakage, which is used to implement an abnormal event warning method for monitoring gas leakage. The system includes a cloud data center, a gas transmission pipeline network, a plurality of gas leakage monitoring devices, and a plurality of gas leakage response devices. The cloud data center is communicatively connected to the plurality of gas leakage monitoring devices and the plurality of gas leakage response devices, respectively. The gas transmission pipeline network is provided with a plurality of monitoring points and a plurality of response points. The plurality of gas leakage monitoring devices are correspondingly provided at the plurality of monitoring points of the gas transmission pipeline network, and the plurality of gas leakage response devices are correspondingly provided at the plurality of response points of the gas transmission pipeline network.

[0126] The cloud data center is used to build a digital twin model for gas leak monitoring using digital twin technology and an abnormal event warning model using artificial intelligence algorithms. Based on multiple real-time gas leak monitoring data collected at the same time, the abnormal event warning model is used to issue abnormal event warnings and obtain real-time abnormal event warning results. Based on the real-time abnormal event warning results, the gas leak monitoring digital twin model is used to visualize abnormal event warnings and control corresponding gas leak response equipment.

[0127] Gas transmission pipeline network for transmitting gas;

[0128] Gas leakage monitoring equipment is used to collect real-time gas leakage monitoring data from monitoring points at each collection moment and upload the real-time gas leakage monitoring data to the cloud data center;

[0129] Gas leak response equipment is used to execute real-time gas leak response instructions and respond to abnormal events.

[0130] The present invention discloses a method and system for monitoring abnormal event warning for gas leakage, applies digital twin technology to gas leakage monitoring, and simulates the actual operating status of the gas transmission pipeline network by constructing a digital twin model for gas leakage monitoring, thereby improving the intuitiveness and convenience of abnormal event warning. An abnormal event warning model is constructed using an artificial intelligence algorithm, including a gas leakage analysis model, a gas leakage positioning model, and an abnormal event response model, thereby realizing intelligent monitoring and early warning of gas leakage, accurately locating the location of the gas leakage, improving the abnormal event warning effect, and providing a response mechanism, providing accurate and appropriate response strategies for the abnormal events warned, thereby improving the intelligence, efficiency and accuracy of gas leakage monitoring; unified data analysis and abnormal event warning visualization are performed based on the cloud data center, thereby improving the real-time performance of gas leakage monitoring and the abnormal event warning response speed.

[0131] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for early warning of abnormal events of monitoring gas leakage, characterized in that: The steps include: Based on the cloud data center, digital twin technology is used to build a digital twin model for gas leakage monitoring, and artificial intelligence algorithms are used to build an abnormal event warning model; Based on the gas leakage monitoring equipment, real-time gas leakage monitoring data of the monitoring point is collected at each collection moment, and the real-time gas leakage monitoring data is uploaded to the cloud data center; Based on the cloud data center, according to a number of real-time gas leakage monitoring data at the same collection time, the abnormal event warning model is used to carry out abnormal event warning and obtain real-time abnormal event warning results; Based on the cloud data center, according to the real-time abnormal event warning results, the gas leakage monitoring digital twin model is used to visualize the abnormal event warning and control the corresponding gas leakage response equipment.

2. The method for early warning of abnormal events for monitoring gas leakage according to claim 1, characterized in that: The abnormal event warning model includes a gas leakage analysis model, a gas leakage location model and an abnormal event response model; The real-time abnormal event warning results include real-time gas leakage analysis results, real-time gas leakage location results and real-time abnormal event response strategies; The real-time abnormal event response strategy includes a real-time abnormal event warning signal and a real-time gas leakage response instruction of a gas leakage response device.

3. The method for early warning of abnormal events for monitoring gas leakage according to claim 2, characterized in that: Based on the cloud data center, digital twin technology is used to build a gas leakage monitoring digital twin model, and artificial intelligence algorithms are used to build an abnormal event warning model, including the following steps: Based on the cloud data center and the topological information of the gas transmission pipeline network, a static sub-model of the gas transmission pipeline network topology is constructed using digital twin technology; According to the first basic attribute information of all gas leakage monitoring equipment, using digital twin technology, a corresponding dynamic sub-model of the gas leakage monitoring equipment is set at the corresponding position of the static sub-model of the gas transmission pipeline network topology; According to the second basic attribute information of all gas leakage response equipment, the digital twin technology is used to set the corresponding gas leakage response equipment dynamic sub-model at the corresponding position of the gas transmission pipeline network topology static sub-model; Integrate the static sub-model of the gas transmission pipeline network topology, the dynamic sub-models of several gas leakage monitoring devices, and the dynamic sub-models of several gas leakage response devices to obtain a digital twin model for gas leakage monitoring. Based on some historical gas leakage monitoring data, a deep learning algorithm is used to build a gas leakage analysis model and generate some historical gas leakage analysis results; Based on the historical gas leakage analysis results, a number of historical gas leakage monitoring data and a gas leakage monitoring digital twin model with gas leakage were used to build a gas leakage positioning model using a deep learning algorithm, and generate a number of historical gas leakage positioning results; Based on several historical gas leakage analysis results and the corresponding historical gas leakage location results, a reinforcement learning algorithm is used to build an abnormal event response model and generate several historical abnormal event response experiences.

4. The method for early warning of abnormal events for monitoring gas leakage according to claim 3 is characterized in that: The gas leakage analysis model is constructed based on the RF-LSTM algorithm; The gas leakage location model is constructed based on the GNN-Attention-LSTM algorithm; The abnormal event response model is constructed based on the DQN algorithm.

5. The method for early warning of abnormal events for monitoring gas leakage according to claim 4, characterized in that: Based on the cloud data center, according to a number of real-time gas leakage monitoring data at the same collection time, the abnormal event warning model is used to perform abnormal event warning, and the real-time abnormal event warning result is obtained, including the following steps: Based on the cloud data center, according to a number of real-time gas leakage monitoring data at the same collection time, the gas leakage analysis model is used to perform gas leakage analysis and obtain real-time gas leakage analysis results; If the real-time gas leakage analysis result shows that there is a gas leakage, then proceed to the next step; otherwise, wait for some real-time gas leakage monitoring data at the next collection time; Based on a number of real-time gas leakage monitoring data and a gas leakage monitoring digital twin model, a gas leakage positioning model is used to locate the gas leakage and obtain a real-time gas leakage positioning result; According to the real-time gas leakage analysis results and the corresponding real-time gas leakage positioning results, an abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy.

6. The method for early warning of abnormal events of monitoring gas leakage according to claim 5, characterized in that: Based on the cloud data center, according to a number of real-time gas leakage monitoring data at the same collection time, a gas leakage analysis model is used to perform gas leakage analysis to obtain real-time gas leakage analysis results, including the following steps: Based on the cloud data center, several real-time gas leakage monitoring data at the same collection time are converted into a real-time gas leakage monitoring data matrix; Using the gas leakage analysis model, M real-time key features of the real-time gas leakage monitoring data matrix are extracted, where M is the total number of key features; According to the M real-time key features, gas leakage analysis is performed to obtain real-time gas leakage analysis results.

7. The method for early warning of abnormal events of monitoring gas leakage according to claim 6, characterized in that: According to a number of real-time gas leakage monitoring data and a gas leakage monitoring digital twin model, a gas leakage positioning model is used to locate the gas leakage and obtain a real-time gas leakage positioning result, including the following steps: Using the gas leakage location model, the real-time gas transmission pipeline network topology characteristics of the static sub-model of the gas transmission pipeline network topology in the gas leakage monitoring digital twin model are extracted; According to the preset attention weights, the M real-time key features of the real-time gas leakage monitoring data matrix and the real-time gas transmission pipeline network topology features are spliced ​​to obtain real-time splicing features; According to the real-time splicing features, the gas leakage is located and the real-time gas leakage location result is obtained.

8. The method for early warning of abnormal events of monitoring gas leakage according to claim 7, characterized in that: According to the real-time gas leakage analysis results and the corresponding real-time gas leakage positioning results, the abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy, including the following steps: According to the real-time gas leakage analysis result and the corresponding real-time gas leakage location result, a search is performed in several historical abnormal event response experiences to obtain several matching historical abnormal event response experiences; According to the real-time gas leakage analysis result, the real-time gas leakage location result and several matching historical abnormal event response experiences, the model parameters of the abnormal event response model are updated to obtain updated model parameters; According to the updated model parameters, the abnormal event response model is used to generate an abnormal event response strategy to obtain a real-time abnormal event response strategy.

9. The method for early warning of abnormal events of monitoring gas leakage according to claim 8, characterized in that: Based on the cloud data center, according to the real-time abnormal event warning results, the gas leakage monitoring digital twin model is used to visualize the abnormal event warning and control the corresponding gas leakage response equipment, including the following steps: Based on the cloud data center, according to the real-time gas leakage location results of the existing gas leakage, the real-time leakage point is located and highlighted in the static sub-model of the gas transmission pipeline network topology of the gas leakage monitoring digital twin model; Add and visualize the real-time gas leakage analysis results and the real-time abnormal event warning signals of the real-time abnormal event response strategy in the dynamic sub-model of the gas leakage monitoring equipment closest to the real-time leakage point; Add and visualize the corresponding real-time gas leakage response instructions in the dynamic sub-models of all gas leakage response equipment involved in the real-time abnormal event response strategy; The real-time gas leakage response instruction is sent to the gas leakage response device at the corresponding response point, and the corresponding gas leakage response device is controlled according to the real-time gas leakage response instruction.

10. An abnormal event early warning system for monitoring gas leakage, used to implement the abnormal event early warning method for monitoring gas leakage as claimed in any one of claims 1 to 9, characterized in that: The system includes a cloud data center, a gas transmission pipeline network, a plurality of gas leakage monitoring devices and a plurality of gas leakage response devices. The cloud data center is communicatively connected with the plurality of gas leakage monitoring devices and the plurality of gas leakage response devices respectively. The gas transmission pipeline network is provided with a plurality of monitoring points and a plurality of response points. The plurality of gas leakage monitoring devices are correspondingly arranged at the plurality of monitoring points of the gas transmission pipeline network, and the plurality of gas leakage response devices are correspondingly arranged at the plurality of response points of the gas transmission pipeline network.

Images