Valve well monitoring and early warning method and system

Abstract

The invention is suitable for the field of valve well monitoring, and provides a valve well monitoring and early warning method and system, a multi-mode composite sensor group is integrated in a valve well, and multi-dimensional real-time data in the valve well is collected; based on signal intensities of different areas in the valve well, an optimal transmission path is matched, and multi-dimensional real-time data is sent; importing a risk intelligent analysis model constructed based on a deep learning neural network algorithm, and generating valve well advanced early warning information; matching and triggering an early warning mechanism with a hierarchical linkage function based on the advanced early warning information of the valve well; acquiring early warning mechanism information, detecting local network connection information of the valve well, and executing a safety strategy of the associated control valve in the valve well in a target time period; the multi-modal composite sensor group realizes comprehensive acquisition of data, and optimal transmission path matching ensures timely transmission of the data, thereby providing guarantee for subsequent analysis. Potential risks are pre-warned, and a hierarchical linkage pre-warning mechanism can make corresponding responses for different risks.

Description

Technical Field

[0001] The present invention belongs to the field of valve well monitoring, and particularly relates to a valve well monitoring and early warning method and system. Background Technique

[0002] With the popularization of urban gas supply, urban gas pipelines, as important infrastructure for urban gas transmission, are widely distributed in areas such as residential areas, industrial parks, and farmlands. These areas have frequent human activities, and farm machinery operations are also relatively common. As a key node of the pipeline, the valve well plays an important role in controlling and regulating gas transmission, and is mostly located near users and key positions of the pipeline.

[0003] However, there are many potential safety hazards in current valve wells. On the one hand, due to long-term influence of external environmental factors such as humidity and corrosion, gas pipelines and related equipment in the valve well are prone to leakage. Once gas leakage occurs, in densely populated residential areas or industrial parks, it is extremely easy to trigger serious accidents such as explosions and fires, posing a huge threat to the safety of residents' lives and property. On the other hand, in areas such as farmlands, frequently operating farm machinery may accidentally touch the valve well facilities, causing mechanical damage and then leading to gas leakage. At the same time, backflow of rainwater and groundwater may also cause the valve well to flood, affecting the normal operation of the equipment in the valve well and reducing its service life.

[0004] Existing valve well monitoring methods are relatively backward, mostly relying on manual regular inspections, which are not only inefficient but also difficult to achieve real-time monitoring, and it is impossible to detect and handle potential risks in a timely manner. Therefore, it is of great practical significance to develop a valve well monitoring and early warning method that can detect valve well risks in real time and effectively prevent and dispose of risk situations. Summary of the Invention

[0005] The purpose of the present invention is to provide a valve well monitoring and early warning method and system, aiming to solve the problems raised in the above background technique.

[0006] The present invention is implemented as follows. On the one hand, a valve well monitoring and early warning method, the method includes: Integrate a multi-modal composite sensor group in the valve well to collect multi-dimensional real-time data in the valve well. The multi-modal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; Match the optimal transmission path based on the signal strength in different regions of the valve well and send multi-dimensional real-time data; Import the multi-dimensional real-time data into a risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate valve well early warning information in advance; Based on the valve well early warning information in advance, match and trigger an early warning mechanism with a hierarchical linkage function; Obtain early warning mechanism information, detect the local network connection information of the valve well, and execute the safety strategy of the associated control valve in the valve well during the target period.

[0007] As a further solution of the present invention, the matching of the optimal transmission path based on the signal strength in different regions of the valve well and the sending of multi-dimensional real-time data specifically include: Obtain the real-time intensity information and real-time bandwidth information of a number of preset signal detection points; Obtain the preset transmission path database; Based on the real... of a number of preset signal detection points.

[0008] As a further solution of the present invention, the importing of multi-dimensional real-time data into the risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate early warning information for the valve well specifically includes: Extract the target feature information from the multi-dimensional real-time data; Import the target feature information into the risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate the early warning level and type; The construction process of the risk intelligent analysis model is as follows: Select the recurrent neural network and its variant long short-term memory network as the network architecture, divide the multi-dimensional real-time data collected in the valve well into a training set, a validation set and a test set, use the training set to train the built model, continuously adjust the weights and biases of the model through the backpropagation algorithm, use the validation set to evaluate the performance of the model, adjust the model parameters or training strategy according to the evaluation results of the validation set, and use the test set to evaluate the trained model.

[0009] As a further solution of the present invention, the matching and triggering of the early warning mechanism with hierarchical linkage function based on the early warning information of the valve well specifically includes: Obtain and analyze the early warning level and type; Match the early warning strategy corresponding to the early warning level and type in the pre-stored early warning mechanism database; Based on the early warning strategy, send the associated personnel notification information and the associated device monitoring instruction.

[0010] As a further solution of the present invention, the obtaining of the early warning mechanism information, the detection of the local network connection information of the valve well, and the execution of the safety strategy of the associated control valve in the valve well during the target period specifically include: Judge whether the early warning strategy is a non-low-level early warning strategy; If the early warning strategy is a non-low-level early warning strategy, monitor whether the associated device is closed during the target period; If it is monitored that the associated device is not closed during the target period, generate an automatic closing instruction for the associated device.

[0011] As a further aspect of the present invention, on the other hand, a valve well monitoring and early warning system, the system comprising: An acquisition module, configured to integrate a multi-modal composite sensor group in the valve well to acquire multi-dimensional real-time data in the valve well; The multi-modal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; A matching module, configured to match an optimal transmission path based on signal strengths in different regions in the valve well; A sending module, configured to send multi-dimensional real-time data; A first import module, configured to import multi-dimensional real-time data into a risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generation module, configured to generate early warning information for the valve well in advance; A matching and triggering module, configured to match and trigger an early warning mechanism with a hierarchical linkage function based on the early warning information for the valve well in advance; An acquisition module, configured to acquire early warning mechanism information; A detection module, configured to detect valve well local network connection information; An execution module, configured to execute the safety strategy of the associated control valve in the valve well.

[0012] As a further aspect of the present invention, the matching module specifically includes: A first acquisition unit, configured to acquire real-time intensity information and real-time bandwidth information of a plurality of preset signal detection points; A second acquisition unit, configured to acquire a preset transmission path database; A calculation unit, configured to calculate a real-time optimal transmission path based on the real-time intensity information, real-time bandwidth information of a plurality of preset signal detection points, and the preset transmission path database, using a genetic algorithm.

[0013] As a further aspect of the present invention, the generation module specifically includes: An extraction unit, configured to extract target feature information from the multi-dimensional real-time data; A second import unit, configured to import the target feature information into a risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generation unit, configured to generate an early warning level and type in advance.

[0014] As a further aspect of the present invention, the matching and triggering module specifically includes: An acquisition and analysis unit, configured to acquire and analyze the early warning level and type in advance; A matching unit, configured to match the early warning level and type with the corresponding early warning strategy in a pre-stored early warning mechanism database; An occurrence unit, configured to send notification information of associated personnel and monitoring instructions for associated devices based on an early warning strategy.

[0015] A valve well monitoring and early warning method and system provided by the present invention. The multi-modal composite sensor group of this method and system realizes comprehensive data collection, and the optimal transmission path matching ensures timely data transmission, providing guarantee for subsequent analysis. Early warning of potential risks, and the hierarchical linkage early warning mechanism can make corresponding responses to different risks. The associated control valve safety strategy can ensure the safety of the valve well at critical moments, avoid manual monitoring errors, and reduce the probability of accidents. Description of the Drawings

[0016] Figure 1 is the main flow chart of a valve well monitoring and early warning method.

[0017] Figure 2 is the flow chart of sending multi-dimensional real-time data by matching the optimal transmission path based on the signal strength in different areas within the valve well in a valve well monitoring and early warning method.

[0018] Figure 3 is the flow chart of importing multi-dimensional real-time data into a risk intelligent analysis model constructed based on a deep learning neural network algorithm to generate early warning information for the valve well in a valve well monitoring and early warning method.

[0019] Figure 4 is the flow chart of matching and triggering an early warning mechanism with a hierarchical linkage function based on the early warning information of the valve well in a valve well monitoring and early warning method.

[0020] Figure 5 is the flow chart of obtaining early warning mechanism information, detecting the local network connection information of the valve well, and executing the safety strategy of the associated control valve in the valve well within the target time period in a valve well monitoring and early warning method.

[0021] Figure 6 is the main structure diagram of a valve well monitoring and early warning system.

[0022] Figure 7 is the structural block diagram of the matching module in a valve well monitoring and early warning system.

[0023] Figure 8 is the structural block diagram of the generating module in a valve well monitoring and early warning system.

[0024] Figure 9 is the structural block diagram of the matching and triggering module in a valve well monitoring and early warning system. Detailed Embodiments

[0025] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0026] The following describes in detail the specific implementation of the present invention with reference to specific embodiments.

[0027] A valve well monitoring and early warning method and system provided by the present invention solve the technical problems in the background art.

[0028] As Figure 1 shown, it is a main flow chart of a valve well monitoring and early warning method provided by an embodiment of the present invention. The valve well monitoring and early warning method includes: Step S100: Integrate a multi-modal composite sensor group in the valve well to collect multi-dimensional real-time data in the valve well; The multi-modal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; Step S200: Based on the signal strength in different regions of the valve well, match the optimal transmission path and send multi-dimensional real-time data; Step S300: Import the multi-dimensional real-time data into a risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate valve well early warning information; Step S400: Based on the valve well early warning information, match and trigger an early warning mechanism with a hierarchical linkage function; Step S500: Obtain the early warning mechanism information, detect the valve well local network connection information, and execute the safety strategy of the associated control valve in the valve well during the target time period; When this embodiment is applied, the gas leakage sensor deployed in the valve well can accurately detect extremely small amounts of gas leakage using advanced gas molecule recognition technology. Once an abnormality is found, data collection is immediately started; the displacement sensor uses the principle of electromagnetic induction to monitor the tiny displacement changes of the valve well structure in real time, providing a key basis for judging the stability of the well body; the water level sensor uses pressure sensing technology to accurately measure the water level in the valve well and promptly detect possible hidden dangers of water accumulation; the stress and strain sensor is based on the principle of resistance strain to sense the stress and strain state of each part of the valve well to prevent safety problems caused by uneven structural force; the temperature and humidity coupling sensor uses sensitive elements to synchronously obtain the temperature and humidity data in the valve well, comprehensively evaluate the impact of environmental factors on the equipment, and then based on the difference in signal strength in different areas of the valve well, the system uses an intelligent signal analysis algorithm to automatically match the optimal transmission path. By dynamically adjusting the data transmission link, signal interference and attenuation are effectively avoided, ensuring that multi-dimensional real-time data can be efficiently and stably transmitted to the data processing center, and importing a risk intelligence analysis model based on a deep learning neural network algorithm. The model is trained based on a large amount of historical data and real-time monitoring data, and can deeply explore the potential connections between data. Through the analysis and processing of multi-dimensional real-time data, potential risks can be accurately identified and advance warning information for valve wells can be generated in advance. Once the warning information is obtained, the system will match and trigger the warning mechanism with hierarchical linkage function according to the risk level. At the same time, the system obtains the warning mechanism information in real time, detects the local network connection information of the valve well, and executes the safety strategy of the associated control valve in the valve well according to the preset strategy during the target period, so as to achieve effective risk control and ensure the safe and stable operation of the valve well.

[0029] like Figure 2 As shown, as a preferred embodiment of the present invention, the method of matching the optimal transmission path based on the signal strength of different areas in the valve well and sending multi-dimensional real-time data specifically includes: Step S201: obtaining real-time strength information and real-time bandwidth information of a plurality of preset signal detection points; Step S202: obtaining a preset transmission path database; Step S203: Based on the real-time strength information of several preset signal detection points, the real-time bandwidth information and the preset transmission path database, a genetic algorithm is used to calculate the real-time optimal transmission path; When this embodiment is applied, first, real-time intensity information and real-time bandwidth information of several preset signal detection points are obtained through signal detection devices reasonably deployed in the transmission network. The real-time intensity information and real-time bandwidth information reflect the signal quality and the data transmission volume that can be carried at each point in the current network, and are important bases for subsequent path calculation. At the same time, a preset transmission path database is obtained. This preset database stores all possible transmission paths and their related parameters in the network, such as path length, node connection relationship, historical transmission performance, etc., providing rich basic data for the genetic algorithm. Based on the obtained real-time intensity information, real-time bandwidth information, and the preset transmission path database, the genetic algorithm is used to calculate the real-time optimal transmission path. The genetic algorithm simulates natural selection and genetic mechanisms, represents the transmission path as a chromosome, and searches among a large number of possible path combinations through operations such as selection, crossover, and mutation, and continuously iterates.

[0030] As Figure 3 shown, as a preferred embodiment of the present invention, importing the multi-dimensional real-time data into the risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate the valve well early warning information specifically includes: Step S301: Extract the target feature information from the multi-dimensional real-time data; Step S302: Import the target feature information into the risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate the early warning level and type; The construction process of the risk intelligent analysis model is as follows: Select the recurrent neural network and its variant long short-term memory network as the network architecture, divide the multi-dimensional real-time data collected in the valve well into a training set, a validation set, and a test set, use the training set to train the built model, continuously adjust the weights and biases of the model through the backpropagation algorithm, use the validation set to evaluate the performance of the model, adjust the model parameters or training strategy according to the evaluation results of the validation set, and use the test set to evaluate the trained model; When this embodiment is applied, the multi-modal composite sensor group in the valve well continuously collects multi-dimensional real-time data such as gas leakage, displacement, water level, stress and strain, and temperature and humidity. Through a preset data processing algorithm, targeted features closely related to the risks of the valve well are screened out from these data. For example, features such as the change trend of the leakage rate and the concentration peak are extracted from the gas leakage data; the frequency and amplitude of displacement mutations are obtained from the displacement data. Signal processing technologies such as wavelet transform and Fourier transform, as well as feature selection methods based on statistics, are used to ensure that the extracted targeted features can accurately reflect the potential risk status of the valve well. Then, the targeted feature information extracted is imported into a risk intelligent analysis model constructed based on the deep learning neural network algorithm. This model has been trained with a large amount of historical data and can perform in-depth analysis on the input targeted features. According to the preset risk assessment logic and combined with the parameters obtained from deep learning training, the model outputs the early warning level and type. The early warning level can be divided into different levels such as low, medium, and high, and the early warning types cover gas leakage risks, structural displacement risks, water accumulation risks, etc.

[0031] As Figure 4 shown, as a preferred embodiment of the present invention, the matching and triggering of an early warning mechanism with a hierarchical linkage function based on the valve well early warning information specifically includes: Step S401: Obtain and analyze the early warning level and type; Step S402: Match the early warning strategies corresponding to the early warning level and type in the pre-stored early warning mechanism database; Step S403: Based on the early warning strategy, send associated personnel notification information and associated device monitoring instructions; It should be understood that an early warning mechanism database is established in advance. This database stores a variety of early warning mechanism strategies formulated for different early warning levels and types. First, obtain and analyze the early warning level and type, and match the early warning strategies corresponding to the early warning level and type in the pre-stored early warning mechanism database. For example, if a high-level gas leakage risk early warning is received, the system will match the early warning mechanism specifically for large-scale gas leakage. If it is a low-level structural displacement risk early warning, a relatively mild response strategy will be matched, such as arranging professional personnel to conduct on-site investigation and evaluation. Based on the early warning strategy, send associated personnel notification information and associated device monitoring instructions. In the early warning mechanism database, information of various responsible personnel is stored. When the specific early warning strategy is confirmed, information will be sent to the relevant associated responsible persons for notification, and at the same time, the status monitoring of the corresponding associated devices will be carried out.

[0032] As Figure 5 shown, as a preferred embodiment of the present invention, the obtaining of the early warning mechanism information, detecting the valve well local network connection information, and executing the safety strategy of the associated control valve in the valve well within the target time period specifically includes: Step S501: Determine whether the early warning strategy is a non-low-level early warning strategy; Step S502: If the early warning strategy is a non-low-level early warning strategy, monitor whether the associated device is turned off during the target period; Step S503: If the associated device is not turned off during the target period, generate an automatic shutdown command for the associated device; When this embodiment is applied, it is determined whether the early warning strategy is a non-low-level early warning strategy. If it is a non-low-level early warning strategy, it indicates that the risk coefficient is relatively high and needs to be processed in a timely manner. During the target period, the system will monitor the status of the associated device in real time. If the associated device is not turned off during the target period, it indicates that the associated device has been manually turned off, and there may be a risk of human negligence. An automatic shutdown command for the associated device is automatically generated to turn off the associated device.

[0033] As Figure 6 shown, as another preferred embodiment of the present invention, on the other hand, a valve well monitoring and early warning system, the system includes: An acquisition module 100, which is used to integrate a multi-modal composite sensor group in the valve well to acquire multi-dimensional real-time data in the valve well; The multi-modal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; A matching module 200, which is used to match the optimal transmission path based on the signal strength in different regions of the valve well; A sending module 300, which is used to send multi-dimensional real-time data; A first import module 400, which is used to import multi-dimensional real-time data into a risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generation module 500, which is used to generate valve well early warning information in advance; A matching trigger module 600, which is used to match and trigger an early warning mechanism with a hierarchical linkage function based on the valve well early warning information in advance; An acquisition module 700, which is used to acquire early warning mechanism information; A detection module 800, which is used to detect the valve well local network connection information; An execution module 900, which is used to execute the safety strategy of the associated control valve in the valve well; When this embodiment is applied, a multi-modal composite sensor group is integrated in the valve well. The acquisition module 100 acquires multi-dimensional real-time data in the valve well. Based on the signal strengths in different regions of the valve well, the matching module 200 matches the optimal transmission path. The sending module 300 sends the multi-dimensional real-time data. The first import module 400 imports the multi-dimensional real-time data into a risk intelligent analysis model constructed based on a deep learning neural network algorithm. The generating module 500 generates early warning information for the valve well. Based on the early warning information for the valve well, the matching trigger module 600 matches and triggers an early warning mechanism with a hierarchical linkage function. The obtaining module 700 obtains the early warning mechanism information. The detection module 800 detects the local network connection information of the valve well. The execution module 900 executes the safety strategy of the associated control valve in the valve well.

[0034] As Figure 7 shown, as another preferred embodiment of the present invention, the matching module 200 specifically includes: A first obtaining unit 201, configured to obtain the real-time intensity information and real-time bandwidth information of a plurality of preset signal detection points; A second obtaining unit 202, configured to obtain a preset transmission path database; A calculation unit 203, configured to calculate the real-time optimal transmission path by using a genetic algorithm based on the real-time intensity information, real-time bandwidth information of a plurality of preset signal detection points, and the preset transmission path database.

[0035] When this embodiment is applied, the first obtaining unit 201 obtains the real-time intensity information and real-time bandwidth information of a plurality of preset signal detection points, the second obtaining unit 202 obtains the preset transmission path database, and based on the real-time intensity information, real-time bandwidth information of a plurality of preset signal detection points, and the preset transmission path database, the calculation unit 203 calculates the real-time optimal transmission path.

[0036] As Figure 8 shown, as another preferred embodiment of the present invention, the generating module 500 specifically includes: An extraction unit 501, configured to extract the target feature information from the multi-dimensional real-time data; A second import unit 502, configured to import the target feature information into a risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generating unit 503, configured to generate the early warning level and type.

[0037] When this embodiment is applied, the extraction unit 501 extracts the target feature information from the multi-dimensional real-time data, the second import unit 502 imports the target feature information into a risk intelligent analysis model constructed based on a deep learning neural network algorithm, and the generating unit 503 generates the early warning level and type.

[0038] AsFigure 9 As shown in the figure, as another preferred embodiment of the present invention, the matching trigger module 600 specifically includes: An acquisition and parsing unit 601, configured to acquire and parse the early warning level and type; A matching unit 602, configured to match the early warning corresponding to the early warning level and type in the pre-stored early warning mechanism database; An occurrence unit 603, configured to send associated personnel notification information and associated device monitoring instructions based on the early warning strategy; When this embodiment is applied, the acquisition and parsing unit 601 acquires and parses the early warning level and type, the matching unit 602 matches the early warning strategy corresponding to the early warning level and type in the pre-stored early warning mechanism database, and based on the early warning strategy, the occurrence unit 603 sends associated personnel notification information and associated device monitoring instructions.

[0039] The above-mentioned embodiment of the present invention provides a valve well monitoring and early warning method, and provides a valve well monitoring and early warning system. The gas leakage sensor deployed in the valve well uses advanced gas molecule recognition technology to accurately detect extremely small amounts of gas leakage. Once an abnormality is found, data collection is immediately started; the displacement sensor uses the principle of electromagnetic induction to monitor the tiny displacement changes of the valve well structure in real time, providing a key basis for judging the stability of the well body; the water level sensor uses pressure sensing technology to accurately measure the water level in the valve well, and promptly detect possible hidden dangers of water accumulation; the stress and strain sensor is based on the principle of resistance strain to sense the stress and strain state of various parts of the valve well, and prevent safety problems caused by uneven structural force; the temperature and humidity coupling sensor uses sensitive elements to synchronously obtain temperature and humidity data in the valve well, and comprehensively evaluates the impact of environmental factors on the equipment. Then, based on the difference in signal strength in different areas of the valve well, the system uses an intelligent signal analysis algorithm to automatically match the optimal transmission path. By dynamically adjusting the data transmission link, signal interference and attenuation can be effectively avoided, ensuring that multi-dimensional real-time data can be efficiently and stably transmitted to the data processing center, and importing a risk intelligence analysis model based on a deep learning neural network algorithm. The model is trained based on a large amount of historical data and real-time monitoring data, and can deeply explore the potential connections between data. Through the analysis and processing of multi-dimensional real-time data, potential risks can be accurately identified, and advance warning information for valve wells can be generated in advance. Once the warning information is obtained, the system will match and trigger the warning mechanism with hierarchical linkage function according to the risk level. At the same time, the system obtains the warning mechanism information in real time, detects the local network connection information of the valve well, and executes the associated control valve safety strategy in the valve well according to the preset strategy within the target period, so as to achieve effective risk control and ensure the safe and stable operation of the valve well; the multi-modal composite sensor group of this method and system realizes comprehensive data collection, and the optimal transmission path matching ensures timely data transmission, providing guarantee for subsequent analysis. Warning potential risks, the hierarchical linkage warning mechanism can respond accordingly to different risks. The associated control valve safety strategy can ensure the safety of the valve well at critical moments, avoid manual monitoring errors, and reduce the probability of accidents.

[0040] In order to load the above-mentioned method and system and enable it to run smoothly, the system, in addition to the various modules mentioned above, may also include more or fewer components than described above, or a combination of certain components, or different components, for example, it may include input and output devices, network access devices, buses, processors and memories, etc.

[0041] The so-called processor may be a central processing unit, or may also be other general-purpose processors, digital signal processors, application-specific integrated circuits, off-the-shelf programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The above-mentioned processor is the control center of the above-mentioned system, and connects each part through various interfaces and circuits.

[0042] The technical features of the above-mentioned embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0043] The above-mentioned embodiments only represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

[0044] The above is only the preferred embodiment of the present invention, and it is not used to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should all be included in the protection scope of the present invention.

Claims

1. A valve well monitoring and early warning method, characterized in that The method includes: Integrating a multi-modal composite sensor group in the valve well to collect multi-dimensional real-time data in the valve well. The multi-modal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; Based on the signal strengths in different areas of the valve well, matching the optimal transmission path and sending the multi-dimensional real-time data; Importing the multi-dimensional real-time data into a risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate early warning information for the valve well; Based on the early warning information for the valve well, matching and triggering an early warning mechanism with a hierarchical linkage function; Obtaining the early warning mechanism information, detecting the local network connection information of the valve well, and implementing the safety strategy for the associated control valves in the valve well during the target time period.

2. The valve well monitoring and early warning method according to claim 1, characterized in that The specific process of matching the optimal transmission path based on the signal strengths in different areas of the valve well and sending the multi-dimensional real-time data includes: Obtaining the real-time strength information and real-time bandwidth information of several preset signal detection points; Obtaining a preset transmission path database; Based on the real-time strength information, real-time bandwidth information of several preset signal detection points and the preset transmission path database, using the genetic algorithm to calculate the real-time optimal transmission path.

3. The valve well monitoring and early warning method according to claim 1, wherein, The specific process of importing the multi-dimensional real-time data into a risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate early warning information for the valve well includes: Extracting the target feature information from the multi-dimensional real-time data; Importing the target feature information into a risk intelligent analysis model constructed based on the deep learning neural network algorithm to generate the early warning level and type; The construction process of the risk intelligent analysis model is: Selecting a recurrent neural network and its variant long short-term memory network as the network architecture, dividing the multi-dimensional real-time data collected in the valve well into a training set, a validation set, and a test set, using the training set to train the built model, continuously adjusting the weights and biases of the model through the backpropagation algorithm, using the validation set to evaluate the performance of the model, adjusting the model parameters or training strategy according to the evaluation results of the validation set, and using the test set to evaluate the trained model.

4. The valve well monitoring and early warning method according to claim 3, characterized in that, The specific process of matching and triggering an early warning mechanism with a hierarchical linkage function based on the early warning information for the valve well includes: Obtaining and parsing the early warning level and type; Matching the early warning level and type with the corresponding early warning strategies in the pre-stored early warning mechanism database; Based on the early warning strategies, sending notification information to associated personnel and monitoring instructions for associated devices.

5. The valve well monitoring and early warning method according to claim 4, wherein, The specific process of obtaining the early warning mechanism information, detecting the local network connection information of the valve well, and implementing the safety strategy for the associated control valves in the valve well during the target time period includes: Judging whether the early warning strategy is a non-low-level early warning strategy; If the early warning strategy is a non-low-level early warning strategy, monitoring whether the associated devices are closed during the target time period; If it is monitored that the associated devices are not closed during the target time period, generating an automatic closing instruction for the associated devices.

6. A valve well monitoring and early warning system, characterized in that, Applying the valve well monitoring and early warning method according to any one of claims 1-5, the system includes: An acquisition module for integrating a multi-modal composite sensor group in the valve well to collect multi-dimensional real-time data in the valve well; The multimodal composite sensor group includes a gas leakage sensor, a displacement sensor, a water level sensor, a stress and strain sensor, and a temperature and humidity coupling sensor; A matching module, configured to match an optimal transmission path based on the signal strengths in different areas within the valve well; A generating module, configured to send multi-dimensional real-time data; A first importing module, configured to import the multi-dimensional real-time data into a risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generating module, configured to generate early warning information for the valve well; A matching and triggering module, configured to match and trigger an early warning mechanism with a hierarchical linkage function based on the early warning information for the valve well; An obtaining module, configured to obtain early warning mechanism information; A detecting module, configured to detect the local network connection information of the valve well; An executing module, configured to execute the safety strategy of the associated control valve within the valve well.

7. The valve well monitoring and early warning system according to claim 6, characterized in that, The matching module specifically includes: A first obtaining unit, configured to obtain the real-time intensity information and real-time bandwidth information of a plurality of preset signal detection points; A second obtaining unit, configured to obtain a preset transmission path database; A calculating unit, configured to calculate the real-time optimal transmission path by using a genetic algorithm based on the real-time intensity information, real-time bandwidth information of a plurality of preset signal detection points, and the preset transmission path database.

8. The valve well monitoring and early warning system according to claim 6, wherein, The generating module specifically includes: An extracting unit, configured to extract the target feature information from the multi-dimensional real-time data; A second importing unit, configured to import the target feature information into the risk intelligent analysis model constructed based on a deep learning neural network algorithm; A generating unit, configured to generate the early warning level and type.

9. The valve well monitoring and early warning system according to claim 8, wherein The matching and triggering module specifically includes: An obtaining and parsing unit, configured to obtain and parse the early warning level and type; A matching unit, configured to match the early warning strategy corresponding to the early warning level and type in the pre-stored early warning mechanism database; A generating unit, configured to send the associated personnel notification information and the associated device monitoring instruction based on the early warning strategy.

Images