An intelligent monitoring method and system applied to high-medium pressure valves

Abstract

The present invention discloses an intelligent monitoring method and system for high-medium pressure valves. The method includes: collecting data through an embedded multi-sensor network according to the real-time operating status and ambient pressure data of the high-medium pressure valves, and comprehensively processing the collected data by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, adopting an anomaly detection algorithm based on machine learning to real-time identify potential fault modes of the high-medium pressure valves; according to the potential fault modes, generating corresponding maintenance and repair plans based on a decision tree algorithm, and feeding back the maintenance and repair plans to the operators through a mobile application to realize intelligent monitoring of the high-medium pressure valves. By using the embodiments of the present invention, real-time monitoring and fault warning of valves can be realized, which is also helpful to optimize the maintenance and repair plans and improve the intelligent level of valve management.

Description

Technical Field

[0001] The present invention belongs to the technical field of high and medium pressure valves, and particularly relates to an intelligent monitoring method and system for high and medium pressure valves. Background Technique

[0002] In the modern industrial production process, high and medium pressure valves, as important control components, are widely used in fields such as petroleum, natural gas, chemical industry, electric power, and water supply. They play a crucial role in the circulation and control of media, ensuring the safety and efficiency of the production process. Therefore, the safety, stability, and reliability of valves have become important issues that cannot be ignored in industrial production.

[0003] Traditionally, the monitoring and maintenance of high and medium pressure valves mainly rely on manual inspections and regular maintenance. This method not only highly depends on the experience of operators, but also, due to the limitations of the frequency and accuracy of manual inspections, often fails to detect potential valve failures in a timely manner, leading to possible safety accidents and economic losses. In addition, regular maintenance plans often cannot target specific operating conditions and actual needs, which may result in over-maintenance or under-maintenance, further affecting production efficiency and the service life of equipment. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent monitoring method and system for high and medium pressure valves to solve the deficiencies in the prior art. It can utilize an embedded multi-sensor network, an adaptive data fusion algorithm, and machine learning-based anomaly detection means to not only achieve real-time monitoring and fault warning of valves, but also help optimize maintenance and repair plans, and improve the intelligent level of valve management.

[0005] An embodiment of the present application provides an intelligent monitoring method for high and medium pressure valves, and the method includes:

[0006] Collect data through an embedded multi-sensor network according to the real-time operating status and environmental pressure data of high and medium pressure valves, and comprehensively process the collected data using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics, wherein the multi-sensor network at least includes a pressure sensor, a temperature sensor, and a vibration sensor;

[0007] Based on the valve working characteristics, use a machine learning-based anomaly detection algorithm to real-time identify potential fault modes of the high and medium pressure valves;

[0008] According to the potential fault modes, generate corresponding maintenance and repair plans based on a decision tree algorithm, and feedback the maintenance and repair plans to the operator through a mobile application to achieve intelligent monitoring of high and medium pressure valves.

[0009] Optionally, based on the real-time operating status of the high-medium pressure valve and environmental pressure data, data is collected through an embedded multi-sensor network, and the collected data is comprehensively processed using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics, including:

[0010] For the data stream of each sensor, continuous wavelet transform is applied to extract features at different scales, and the features represent the change details and trends of the signal to capture multi-level information of the data;

[0011] A perception game model is constructed, with each sensor regarded as a participant in the game. Each sensor participates in the game according to the contribution degree and reliability of the data it collects, competing for importance in the fusion process. Through iterative calculation, a Nash equilibrium point is reached, at which point the data weight distribution of each sensor reaches an optimal state;

[0012] The features at different scales extracted from different sensors are input into a fuzzy logic system, and a fused feature is output as the comprehensive valve working characteristic. Among them, the fuzzy rule set is adjusted in real time according to the weights assigned by the perception game model to adapt to different operating states and environments, and a fuzzy reasoning process is performed. Through the matching of fuzzy sets and fuzzy rules, the features at different scales of the sensors are transformed into comprehensive fused features, reflecting the working state of the valve.

[0013] Optionally, based on the valve working characteristics, an anomaly detection algorithm based on machine learning is used to real-time identify potential fault modes of the high-medium pressure valve, including:

[0014] The valve working characteristics are input into a deep autoencoder network to learn its low-dimensional representation, and the output of the bottleneck layer of the autoencoder is used as a feature embedding. This embedding represents the core features of the data and can distinguish normal and abnormal modes. Among them, the autoencoder captures the potential structural features of the data through the reconstruction error;

[0015] Based on the feature embedding, the isolation forest algorithm is used for anomaly detection. The isolation forest constructs decision trees by randomly selecting features and split points to identify abnormal data points different from the normal mode;

[0016] A causal inference network is used to analyze the causal relationship between the anomaly detection results of the isolation forest algorithm and the operating conditions of the high-medium pressure valve to identify potential fault modes. Among them, the causal inference network reveals the driving factors behind abnormal phenomena through a causal graph model.

[0017] Optionally, according to the potential fault modes, a corresponding maintenance and repair plan is generated based on the decision tree algorithm, including:

[0018] The potential failure mode is represented as a binary vector using one-hot encoding, and combined with the numerical characteristics of the collected data to form a complete feature vector as the input of the decision tree model;

[0019] Starting from the root node of the decision tree, traverse the tree structure layer by layer. For each node, use the feature division condition of the node to compare the feature vector. Among them, if a certain feature value in the feature vector meets a certain preset condition, continue to traverse downward along the branch that meets the preset condition. If a certain feature value does not meet a certain preset condition, continue to traverse downward along the other branch until reaching the leaf node;

[0020] After path tracing, finally reach a leaf node, which represents specific maintenance and repair suggestions, and generate a complete maintenance and repair plan based on the leaf nodes reached by tracing.

[0021] Another embodiment of the present application provides an intelligent monitoring system applied to high and medium pressure valves. The system includes:

[0022] An acquisition module, configured to collect data through an embedded multi-sensor network according to the real-time operating state and environmental pressure data of the high and medium pressure valves, and comprehensively process the collected data using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics. Among them, the multi-sensor network at least includes a pressure sensor, a temperature sensor, and a vibration sensor;

[0023] An identification module, configured to based on the valve working characteristics, adopt an anomaly detection algorithm based on machine learning to identify the potential failure mode of the high and medium pressure valves in real time;

[0024] A monitoring module, configured to generate a corresponding maintenance and repair plan based on the decision tree algorithm according to the potential failure mode, and feedback the maintenance and repair plan to the operator through a mobile application to realize the intelligent monitoring of the high and medium pressure valves.

[0025] Another embodiment of the present application provides a storage medium, in which a computer program is stored. Among them, the computer program is set to execute the method described in any one of the above when running.

[0026] Another embodiment of the present application provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is set to run the computer program to execute the method described in any one of the above.

[0027] Compared with the prior art, an intelligent monitoring method for high-medium pressure valves provided by the present invention collects data through an embedded multi-sensor network according to the real-time operating state and ambient pressure data of the high-medium pressure valves, and comprehensively processes the collected data by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, an anomaly detection algorithm based on machine learning is adopted to real-time identify potential fault modes of the high-medium pressure valves; according to the potential fault modes, a corresponding maintenance and repair plan is generated based on a decision tree algorithm, and the maintenance and repair plan is fed back to the operator through a mobile application to realize the intelligent monitoring of the high-medium pressure valves, so that by using the embedded multi-sensor network, the adaptive data fusion algorithm and the anomaly detection means based on machine learning, not only can the real-time monitoring and fault warning of the valves be realized, but also it is helpful to optimize the maintenance and repair plan and improve the intelligent level of valve management. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 FIG. is a hardware structure block diagram of a computer terminal for an intelligent monitoring method for high-medium pressure valves provided by an embodiment of the present invention;

[0029] Figure 2 FIG. is a schematic flowchart of an intelligent monitoring method for high-medium pressure valves provided by an embodiment of the present invention;

[0030] Figure 3 FIG. is a schematic structural diagram of an intelligent monitoring system for high-medium pressure valves provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and should not be construed as limiting the present invention.

[0032] An embodiment of the present invention first provides an intelligent monitoring method for high-medium pressure valves, which can be applied to an electronic device, such as a computer terminal, specifically, a general computer, etc.

[0033] The following takes running on a computer terminal as an example to describe it in detail. Figure 1 FIG. is a hardware structure block diagram of a computer terminal for an intelligent monitoring method for high-medium pressure valves provided by an embodiment of the present invention. As Figure 1 shown, the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory.

[0034] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any intelligent monitoring method for high-medium pressure valves.

[0035] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0036] The internal memory provides an environment for the operation of computer programs in non-volatile storage media. When the computer program is executed by the processor, it enables the processor to execute any intelligent monitoring method applied to high and medium pressure valves.

[0037] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 1 The structure shown in [the figure] is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0038] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0039] See Figure 2 , the embodiments of the present invention provide an intelligent monitoring method applied to high and medium pressure valves, which may include the following steps:

[0040] S201, according to the real-time operating state and environmental pressure data of the high and medium pressure valve, collect data through an embedded multi-sensor network, and comprehensively process the collected data using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics, where the multi-sensor network at least includes a pressure sensor, a temperature sensor, and a vibration sensor;

[0041] In this step, the system collects the real-time operating status of high and medium pressure valves and environmental pressure data through an embedded multi-sensor network. The multi-sensor network includes pressure sensors, temperature sensors, and vibration sensors, and these sensors work together to provide comprehensive environmental and operating condition data. The collected data is comprehensively processed through an adaptive data fusion algorithm to obtain the comprehensive working characteristics of the valve. The adaptive data fusion algorithm can dynamically adjust the data processing strategy to adapt to different operating conditions and environmental changes, thus ensuring the accuracy and reliability of the data.

[0042] The significance of this step is that through the collaborative work of the multi-sensor network, the operating status of the valve and environmental conditions can be monitored in real time. Through the application of the adaptive data fusion algorithm, the system can effectively integrate data from different sensors and generate a comprehensive working characteristic of the valve. This characteristic provides reliable basic data support for subsequent fault detection and maintenance plans, ensuring the safe and efficient operation of the valve in a complex environment.

[0043] Specifically, for the data stream of each sensor, continuous wavelet transform is applied to extract features at different scales. These features represent the change details and trends of the signal to capture multi-level information of the data.

[0044] In this step, the system applies continuous wavelet transform (CWT) to the data stream of each sensor to extract multi-scale features of the signal. Wavelet transform is a signal processing technique that can analyze the signal simultaneously in the time and frequency domains to capture its change details and trends. Through this method, the system can identify multi-level information in the data and provide rich features for subsequent data fusion and analysis.

[0045] The significance of applying continuous wavelet transform is that it can extract more detailed and comprehensive feature information from sensor data. These features not only reflect the instantaneous changes of the signal but also reveal its long-term trends, providing a more accurate and reliable basis for subsequent fault detection and diagnosis.

[0046] In the specific implementation, assume that we have a data stream from a pressure sensor. First, the system preprocesses this data stream to remove noise and outliers. Then, continuous wavelet transform is applied to decompose the data into components at different scales. Each component represents the change of the signal in different frequency ranges. By analyzing these components, the system can identify the detail changes (such as short-term fluctuations) and trend changes (such as long-term pressure changes) of the pressure signal. For example, when abnormal fluctuations in a certain frequency range are detected, the system can infer possible mechanical vibrations or external interferences.

[0047] Exemplarily, assume that we use a pressure sensor in an industrial environment to monitor the working state of a valve. The pressure sensor collects data at a high frequency, generating a time series. To extract useful information from it, we first perform simple preprocessing on the data, such as removing noise and outliers. Then, we apply the continuous wavelet transform (CWT) to analyze the sequence data. By selecting an appropriate wavelet basis (such as the Morlet wavelet), the original signal can be decomposed into wavelet coefficients of different frequency components. These coefficients reflect the changes of the signal at different scales. For example, during a normal operation cycle, the wavelet transform may show a stable trend in the low-frequency region, while abnormal oscillations may be shown in some high-frequency regions, and these oscillations may indicate potential faults or external interferences. By analyzing the characteristics at these different scales, we can understand the multi-level information of the signal, providing a basis for subsequent data fusion and analysis.

[0048] Construct a sensing game model, regarding each sensor as a participant in the game. Each sensor participates in the game according to the contribution degree and reliability of the data it collects, competing for importance in the fusion process. Through iterative calculation, a Nash equilibrium point is reached, at which point the data weight distribution of each sensor reaches the optimal state.

[0049] In this step, the system constructs a sensing game model, regarding each sensor as a participant in the game. Each sensor participates in the game according to the contribution degree and reliability of its data, competing for importance in the data fusion process. Through iterative calculation, the system finally reaches a Nash equilibrium point, at which point the data weight distribution of each sensor reaches the optimal state.

[0050] The application of the sensing game model enables the system to dynamically adjust the data weights of each sensor, ensuring that the most reliable and most contributive data is given higher weights during the data fusion process. This dynamic adjustment mechanism improves the accuracy and robustness of data fusion, ensuring that the system can obtain the best comprehensive features under different environments and operating conditions.

[0051] In a specific implementation, assume that we have pressure, temperature, and vibration sensors. First, the system evaluates the contribution degree and reliability of the data of each sensor. For example, the pressure sensor may contribute more when detecting the valve tightness, while the temperature sensor is more important when detecting thermal expansion. Then, the system regards these sensors as participants in the game and constructs a sensing game model. Through iterative calculation, the system adjusts the weight of each sensor until the Nash equilibrium point is reached. At this point, the pressure sensor may obtain a higher weight because its data is more reliable and important in the current environment.

[0052] Exemplarily, in a valve monitoring system, assume we have three sensors: a pressure sensor, a temperature sensor, and a vibration sensor. We need to determine the weight of each sensor in the data fusion process. First, we establish an initial contribution degree and reliability index for each sensor. For example, the pressure sensor may have a high contribution degree in a high-pressure environment, while the temperature sensor may be more reliable when the ambient temperature changes significantly. Then, we construct a perception game model, in which each sensor participates in the game according to its index. Through multiple rounds of iterative calculations, the sensors will adjust their behaviors according to the strategies of other sensors to obtain more weight in the game. Eventually, the game reaches a Nash equilibrium, and the weight distribution of each sensor reaches an optimal state. For example, during the iterative process, the vibration sensor may have a lower weight because its data does not provide significant fault indications during this operation, while the pressure sensor obtains a higher weight because it provides important status information. This dynamic weight allocation mechanism ensures that the system optimally uses the data of each sensor under different conditions.

[0053] Input the different-scale features extracted from different sensors into the fuzzy logic system, and output a fused feature as the comprehensive valve working feature. Among them, the fuzzy rule set is adjusted in real time according to the weights assigned by the perception game model to adapt to different operating states and environments, perform the fuzzy inference process, and through the matching of the fuzzy set and the fuzzy rule, convert the different-scale features of the sensors into a comprehensive fused feature, reflecting the valve working state.

[0054] In this step, the system inputs the multi-scale features extracted from different sensors into the fuzzy logic system. The fuzzy logic system performs the fuzzy inference process through the matching of the fuzzy set and the fuzzy rule, and converts these features into a comprehensive fused feature. The fuzzy rule set is adjusted in real time according to the weights assigned by the perception game model to adapt to different operating states and environments.

[0055] The application of the fuzzy logic system enables the system to handle uncertainty and ambiguity, and generate a comprehensive feature reflecting the valve working state. By adjusting the fuzzy rule set in real time, the system can adapt to different operating conditions and environmental changes, improving the accuracy and flexibility of fault detection and diagnosis.

[0056] In a specific implementation, assume that we have extracted multi-scale features from pressure, temperature, and vibration sensors. The system first inputs these features into a fuzzy logic system. The fuzzy logic system contains a set of fuzzy rules, such as "If the pressure is high and the temperature is high, then the valve may be overheated". These rules are adjusted according to the weights assigned by the perception game model to ensure that the most important features have the greatest impact on the final decision. Through the fuzzy inference process, the system transforms features of different scales into a comprehensive fused feature, reflecting the overall working state of the valve. For example, when it is detected that both the pressure and temperature increase simultaneously, the system may output a high-risk fused feature, indicating that further inspection is required.

[0057] Exemplarily, in this step, we use a fuzzy logic system to integrate features from different sensors. In a specific implementation, assume that we have extracted multi-scale features from pressure, temperature, and vibration sensors and obtained their respective weights based on the perception game model. When designing the fuzzy logic system, we need to define fuzzy sets and fuzzy rules. For example, the defined fuzzy sets include "low pressure", "medium pressure", "high pressure", as well as "low temperature", "high temperature", etc. The fuzzy rules may be similar to "If the pressure is high and the temperature is high, then the risk level is high". According to the weights given by the game model, we adjust the influence of these rules in real time. After inputting the extracted features into the fuzzy logic system, the system performs an inference process according to the set of fuzzy rules. Finally, a comprehensive fused feature is output. For example, at a certain moment, the pressure and temperature features show abnormally high values. The system obtains a high-risk fused feature value by applying the rules "high pressure" and "high temperature", indicating that the valve needs to be inspected and maintained. This process ensures that the comprehensive feature can accurately reflect the current working state of the valve and provides a reliable basis for decision-making.

[0058] S202, based on the working characteristics of the valve, adopt an anomaly detection algorithm based on machine learning to identify potential fault modes of the high and medium pressure valves in real time;

[0059] In this step, the system, based on the working characteristics of the valve, uses an anomaly detection algorithm of machine learning to identify potential fault modes of high and medium pressure valves in real time. Specifically, the system first obtains the comprehensive working characteristics of the valve from the multi-sensor network, and these characteristics reflect the current operating state of the valve. Then, the system uses machine learning algorithms to analyze these characteristics and identify abnormal modes different from the normal operation mode. In this way, the system can detect potential problems before a fault occurs, thereby improving the safety and reliability of the valve.

[0060] The significance of this step is that through the application of machine learning algorithms, the system can automatically identify potential fault modes of the valve, reducing the dependence on manual monitoring. Real-time detection of abnormal patterns helps to detect problems in advance and avoid downtime and losses caused by failures. By promptly identifying and handling potential faults, the system can extend the service life of the valve, improve the overall operating efficiency and safety of the equipment.

[0061] Specifically, the working characteristics of the valve can be input into a deep autoencoder network to learn its low-dimensional representation. The output of the bottleneck layer of the autoencoder is used as a feature embedding, which represents the core features of the data and can distinguish normal and abnormal patterns. Among them, the autoencoder captures the potential structural features of the data through the reconstruction error.

[0062] In this step, the system inputs the working characteristics of the valve into a deep autoencoder network. The autoencoder is an unsupervised learning model that can learn the low-dimensional representation of data by compressing and reconstructing it. Specifically, the input features are compressed to the bottleneck layer through the encoder part, and the output of the bottleneck layer is the feature embedding, representing the core features of the data. Through the reconstruction error, the autoencoder can capture the potential structural features of the data to help distinguish normal and abnormal patterns.

[0063] The significance of using a deep autoencoder is that it can effectively extract the core features of the data, reduce the dimension of the data, and at the same time retain its important information. By learning the low-dimensional representation, the system can more easily identify abnormal patterns because abnormal data usually generates a large reconstruction error during reconstruction. In this way, the system can more accurately detect potential faults and improve the efficiency and accuracy of anomaly detection.

[0064] In a specific implementation, assume that we have a high-dimensional dataset containing pressure, temperature, and vibration characteristics. First, the system inputs these features into a pre-trained deep autoencoder. The structure of the autoencoder includes an encoder and a decoder. The encoder compresses the input data to the bottleneck layer, and the output of the bottleneck layer is the low-dimensional feature embedding. The decoder then attempts to reconstruct the original input from the output of the bottleneck layer. By minimizing the reconstruction error, the autoencoder can learn the potential structural features of the data. Normal data can usually be well reconstructed, while abnormal data will generate a large reconstruction error. By analyzing the output of the bottleneck layer, the system can identify which data points may be abnormal, providing a basis for subsequent anomaly detection.

[0065] Exemplarily, in an industrial site, we monitor the operating status of a high-medium pressure valve and collect data through pressure sensors, temperature sensors, and vibration sensors. Assuming that each sensor collects 800 data points per second for a week, we will have a large amount of high-dimensional data. To solve the problem of high data dimensionality, we decide to use a deep autoencoder to extract the low-dimensional representation of the data.

[0066] First, we organize the data of a whole day into a high-dimensional matrix containing pressure, temperature, and vibration features. Then, this matrix is fed into a deep autoencoder. The autoencoder consists of multiple layers of neural networks, where the encoder part gradually compresses the input data to the bottleneck layer. For example, the encoder can have three layers. The first layer compresses from 2400 dimensions to 1200 dimensions, the second layer compresses from 1200 dimensions to 600 dimensions, and finally the bottleneck layer compresses the data to 100 dimensions.

[0067] The output of the bottleneck layer represents the low-dimensional feature embeddings, which are learned by the trained autoencoder and can effectively retain the core information of the data. Then, the decoder attempts to reconstruct the original input data. By minimizing the error between the input data and the reconstructed data, the network can capture the potential structural features of the data. By comparing the reconstruction error, we can distinguish normal data from abnormal data. Usually, abnormal data will show a higher reconstruction error because it cannot be reconstructed well.

[0068] Based on the said feature embeddings, an isolation forest algorithm is used for anomaly detection. The isolation forest constructs decision trees by randomly selecting features and split points to identify abnormal data points different from the normal pattern;

[0069] In this step, the system uses the isolation forest algorithm to perform anomaly detection on the feature embeddings. The isolation forest is an unsupervised learning algorithm based on a tree structure. It constructs multiple decision trees by randomly selecting features and split points. The path length of each data point in these trees is used to evaluate its degree of abnormality. Usually, abnormal data points will have a shorter path in the trees because they are easier to be isolated.

[0070] The application of the isolation forest algorithm enables the system to efficiently identify abnormal data points. Due to its randomness and tree structure, the isolation forest can quickly process large-scale data sets and has a high detection accuracy for abnormal data points. By identifying abnormal data different from the normal pattern, the system can give early warnings of potential failures in a timely manner, reducing equipment downtime and maintenance costs.

[0071] In a specific implementation, assume that we have obtained low-dimensional feature embeddings from an autoencoder. The system inputs these embeddings into an isolation forest algorithm. The isolation forest consists of multiple randomly constructed decision trees, and each tree divides the dataset by randomly selecting features and split points. For each data point, the system calculates its average path length in all the trees. Generally, normal data points will have longer paths in the trees, while abnormal data points will have shorter paths. By setting a threshold, the system can identify abnormal data points with shorter path lengths. These abnormal points may correspond to potential failure modes of the valve, and the system can further analyze these points to determine the specific causes of the failures.

[0072] Exemplarily, after obtaining the low-dimensional feature embeddings extracted from the autoencoder, we use the isolation forest algorithm for anomaly detection. The isolation forest consists of multiple randomly constructed binary trees, and each tree divides the data by randomly selecting features and cut points.

[0073] Suppose we have obtained the low-dimensional feature embeddings of 5000 data points from the bottleneck layer. To implement the isolation forest, we construct 100 decision trees. In each tree, the data is divided by randomly selecting features and split points. For each data point, we calculate its average path length in all the trees.

[0074] In actual operation, we find that a certain data point has a short path length in multiple trees, which means that this point is more likely to be isolated compared to other data points. We determine a threshold based on the distribution of these path lengths, and data points exceeding this threshold are marked as abnormal. For example, it is determined that the valve has abnormal vibrations during a certain period. This abnormal data point shows significant isolation in the isolation forest, prompting the operator to conduct further inspections and analyses on it.

[0075] Using a causal inference network, analyze the causal relationship between the anomaly detection results of the isolation forest algorithm and the operating conditions of the high and medium pressure valves, and identify potential failure modes. Among them, the causal inference network reveals the driving factors behind the abnormal phenomena through a causal graph model.

[0076] In this step, the system uses a causal inference network to analyze the causal relationship between the abnormal results detected by the isolation forest algorithm and the valve operating conditions. The causal inference network reveals the potential causal relationships in the data through a causal graph model, helping to identify the driving factors behind the abnormal phenomena. Through this analysis, the system can understand the causes of the abnormal patterns more deeply, thereby identifying potential failure modes.

[0077] The application of the causal inference network enables the system to go beyond simple correlation analysis and deeply explore the root causes of abnormal phenomena. By identifying the causal relationships between abnormal data and operating conditions, the system can more accurately identify fault patterns and provide a basis for formulating effective maintenance strategies. This causal analysis helps improve the accuracy and reliability of fault diagnosis.

[0078] In a specific implementation, assume that the Isolation Forest algorithm detects some abnormal data points. The system combines these abnormal points with the operating conditions of the valve (such as pressure, temperature, vibration, etc.) to construct a causal graph model. The causal graph model represents variables and their causal relationships through nodes and directed edges. The system uses the causal inference network to analyze these relationships and identify which operating conditions may be the driving factors of abnormal phenomena. For example, the system may find that when high pressure and high temperature occur simultaneously, the probability of abnormal data points increases significantly, indicating that these conditions may lead to potential failures of the valve. Through this causal analysis, the system can identify specific fault patterns and provide guidance for subsequent maintenance and repair.

[0079] Exemplarily, after identifying the abnormal data points, we need to figure out the root causes of these abnormalities, so a causal inference network is introduced. The causal inference network analyzes the causal relationships between variables by constructing a causal graph model.

[0080] First, we construct a causal graph where nodes represent different sensor readings and operating conditions, such as pressure levels, temperature changes, vibration intensities, etc. Directed edges represent the causal relationships between variables. For example, an edge in the graph may represent the potential impact of temperature changes on pressure.

[0081] Next, we use machine learning techniques and domain knowledge to fill in the relationships and conditional probabilities between variables in the causal graph. At this time, we can input the abnormal data points marked by the Isolation Forest into the causal inference network for reasoning to obtain which operating conditions or combinations may have caused the abnormal phenomena. For example, through causal reasoning, it is found that high pressure plus high temperature is the main factor for the current abnormal vibration. This information will be fed back to the maintenance team to help them focus on the areas more likely to cause problems for detailed inspection and preventive maintenance. This method is more in-depth than simple correlation analysis because it reveals the causal driving factors behind the abnormalities and helps quickly and accurately formulate fault repair plans.

[0082] S203, generate a corresponding maintenance and repair plan based on the decision tree algorithm according to the potential fault pattern, and feedback the maintenance and repair plan to the operator through the mobile application to achieve intelligent monitoring of high and medium pressure valves.

[0083] In an intelligent monitoring system, according to the identified potential fault modes, the system uses the decision tree algorithm to generate corresponding maintenance and repair plans. The decision tree algorithm analyzes the characteristics of the fault modes and automatically generates a series of maintenance steps and repair suggestions. These suggestions are integrated into a detailed plan and fed back to the operators through a mobile application. The operators can view these plans in real time on mobile devices to ensure timely actions to prevent the further development of faults. In this way, the system can not only identify faults but also provide specific solutions to help the operators respond quickly.

[0084] The significance of this step lies in closely integrating fault detection with actual maintenance operations. By automatically generating maintenance and repair plans, the system can significantly improve the efficiency and accuracy of fault handling. Operators no longer need to rely on experience or manual analysis to formulate maintenance plans but can operate based on the scientific suggestions provided by the system. This automated process reduces the possibility of human errors and ensures the timeliness and effectiveness of maintenance work, thus improving the overall reliability and safety of the equipment.

[0085] Specifically, according to the potential fault modes, to generate corresponding maintenance and repair plans based on the decision tree algorithm, one-hot encoding can be used to represent the potential fault modes as binary vectors, and combined with the numerical characteristics of the collected data to form a complete feature vector as the input of the decision tree model;

[0086] In this step, the system first performs one-hot encoding on the identified potential fault modes and converts them into binary vectors. One-hot encoding is a technique for converting categorical data into binary form, and each fault mode is represented as a unique binary vector. Then, the system combines these binary vectors with the numerical characteristics collected by the sensors to form a complete feature vector. This feature vector contains the encoded information of the fault modes and the real-time data characteristics and serves as the input of the decision tree model.

[0087] By encoding the fault modes as binary vectors, the system can process different types of faults in a structured manner. The complete feature vector formed by combining the numerical characteristics ensures that the decision tree model can comprehensively understand the current equipment status. This method not only improves the input quality of the model but also enhances the model's ability to identify different fault modes, ensuring that the generated maintenance plans are more targeted and effective.

[0088] Suppose the system detects three potential failure modes: A, B, and C. Through one-hot encoding, mode A is represented as [1, 0, 0], mode B as [0, 1, 0], and mode C as [0, 0, 1]. Suppose the data collected by the sensor includes pressure, temperature, and vibration, which are [100, 75, 0.5] respectively. For failure mode A, the complete feature vector is [1, 0, 0, 100, 75, 0.5]. This feature vector is input into the decision tree model, and the model analyzes based on these input features to identify the most suitable maintenance and repair strategy.

[0089] Exemplarily, in an industrial valve monitoring system, the system detects three potential failure modes: A, B, and C. One-hot encoding technology is used to process this categorical information. Failure mode A is encoded as [1, 0, 0], mode B is encoded as [0, 1, 0], and mode C is encoded as [0, 0, 1]. Now suppose the data collected by the sensor includes the following numerical features: the pressure is 105 psi, the temperature is 320 Kelvin, and the vibration amplitude is 0.4 m / s². Suppose the currently identified failure mode is B. Combining the one-hot encoding and the sensor data, the formed complete feature vector is [0, 1, 0, 105, 320, 0.4]. This feature vector is used as the input to the decision tree model. This feature vector not only carries the clear failure mode but also combines the multi-dimensional data collected during the monitoring process, providing rich information for subsequent decision-making analysis.

[0090] Starting from the root node of the decision tree, traverse the tree structure layer by layer. For each node, use the feature partitioning condition of the node to compare the feature vector. Among them, if a certain feature value in the feature vector meets a certain preset condition, continue to traverse downward along the branch that meets this preset condition. If a certain feature value does not meet a certain preset condition, then continue to traverse downward along the other branch until reaching the leaf node;

[0091] In this step, the system starts from the root node of the decision tree and traverses the tree structure layer by layer. Each node contains a feature partitioning condition for comparing the input feature vector. The system determines whether the values in the feature vector meet the partitioning condition of the node according to the values in the feature vector. If it meets, continue to traverse downward along the branch corresponding to this condition; if it does not meet, then continue along the other branch. This process continues until reaching the leaf node.

[0092] By traversing the decision tree layer by layer, the system can effectively analyze the information in the feature vector and gradually narrow down the possible fault range. The division conditions of each node help the system make effective decisions in the complex feature space. When finally reaching the leaf node, the system can provide specific maintenance and repair suggestions. This method ensures the accuracy and efficiency of fault analysis and helps operators quickly identify and solve problems.

[0093] Suppose the root node of the decision tree is divided according to the pressure feature, and the condition is "pressure > 90". If the pressure value in the feature vector is 100, which meets the condition, then continue to traverse downward along the "yes" branch. The next node may be divided according to the temperature feature, and the condition is "temperature > 70". The temperature in the feature vector is 75, still meeting the condition, and continue along the "yes" branch. Finally, reach the leaf node, which may indicate "check the valve tightness". Through this layer-by-layer traversal, the system can provide precise maintenance suggestions according to the specific values of the feature vector.

[0094] Taking a simple decision tree structure as an example, assume that the division condition of the root node is "pressure > 100 psi", and the current pressure value represented by the feature vector is 105 psi, which meets the condition. Therefore, the system continues to traverse along the "yes" branch. The next node may be divided based on temperature, and the condition is set as "temperature > 300 Kelvin". The temperature value in the current feature vector is 320 Kelvin, which also meets this condition, so continue to traverse downward along the "yes" branch. Suppose the next node is based on the vibration amplitude, and the condition is "vibration amplitude < 0.5 m / s²", and the current vibration amplitude is 0.4 m / s², meeting the condition, and the system traverses along the "yes" branch until it reaches the leaf node. Finally, the leaf node reached by the system provides specific maintenance suggestions. This process is like screening layer by layer, applying simple judgment conditions in turn to narrow down the possible solutions to the most suitable option, ensuring that the final suggestion is based on the most directly relevant path of the feature vector.

[0095] After path tracing, finally reach a leaf node, which represents specific maintenance and repair suggestions, and generate a complete maintenance and repair plan based on the leaf nodes reached by the tracing.

[0096] During the path tracing process, the system finally reaches a leaf node. Each leaf node represents a specific maintenance and repair suggestion. The system integrates all the suggestions according to the leaf nodes reached by the path tracing and generates a complete maintenance and repair plan. This plan details the steps and precautions that need to be executed to ensure that the operator can perform maintenance according to the suggestions.

[0097] By generating a complete maintenance and repair plan, the system can provide clear operation guidance for operators. Such a plan not only improves the efficiency of maintenance work but also reduces errors caused by uncertainties. Operators can execute maintenance tasks step by step according to the suggestions in the plan to ensure the normal operation and safety of the equipment.

[0098] Suppose that during the path tracing process, the system reaches three leaf nodes, which respectively provide the following suggestions: "Check the valve seal", "Replace the temperature sensor", and "Calibrate the vibration sensor". The system integrates these suggestions into a maintenance plan, detailing the execution order and precautions for each step. For example, the plan may suggest first checking the valve seal to ensure there is no leakage; then replacing the temperature sensor to improve measurement accuracy; and finally calibrating the vibration sensor to ensure its normal operation. In this way, operators can execute step by step according to the plan to ensure the reliability and safety of the equipment.

[0099] Through the above path tracing, assume that the system finally reaches specific leaf nodes, and the outputs of the leaf nodes are "Conduct a valve seal check" and "Record temperature anomaly data". These outputs are not just individual maintenance suggestions but responses to comprehensive information. Operators will first conduct a valve seal check based on these suggestions to determine whether there are leaks or other related problems. Next, they will record the possible impacts of temperature anomalies, which may involve data recording or further analysis to have more comprehensive reference data for adjusting the system or detecting processes in the future. Based on the suggestions of these leaf nodes, the system generates a detailed maintenance and repair plan. The plan may include: specific steps of the inspection procedure, required tools and materials, estimated completion time, and any possible safety precautions. Through comprehensive suggestions, operators can quickly and effectively solve problems, minimize downtime due to failures as much as possible, and improve the availability and safety of the equipment.

[0100] It can be seen that according to the real-time operation status of high and medium pressure valves and ambient pressure data, data is collected through an embedded multi-sensor network, and the collected data is comprehensively processed using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, an anomaly detection algorithm based on machine learning is used to identify potential fault modes of the high and medium pressure valves in real time; according to the potential fault modes, a corresponding maintenance and repair plan is generated based on the decision tree algorithm, and the maintenance and repair plan is fed back to the operator through a mobile application to achieve intelligent monitoring of high and medium pressure valves, so that by using an embedded multi-sensor network, an adaptive data fusion algorithm, and an anomaly detection means based on machine learning, not only can real-time monitoring and fault warning of valves be achieved, but also it helps to optimize the maintenance and repair plan and improve the intelligent level of valve management.

[0101] Another embodiment of the present invention provides an intelligent monitoring system applied to high-medium pressure valves. Refer to Figure 3 , the system may include:

[0102] An acquisition module 301, configured to collect data through an embedded multi-sensor network according to the real-time operating state of the high-medium pressure valve and the environmental pressure data, and comprehensively process the collected data by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics. Among them, the multi-sensor network at least includes a pressure sensor, a temperature sensor, and a vibration sensor;

[0103] An identification module 302, configured to, based on the valve working characteristics, adopt an anomaly detection algorithm based on machine learning to identify the potential fault modes of the high-medium pressure valve in real time;

[0104] A monitoring module 303, configured to generate a corresponding maintenance and repair plan based on the decision tree algorithm according to the potential fault modes, and feedback the maintenance and repair plan to the operator through a mobile application to realize the intelligent monitoring of the high-medium pressure valve.

[0105] It can be seen that according to the real-time operating state of the high-medium pressure valve and the environmental pressure data, data is collected through an embedded multi-sensor network, and the collected data is comprehensively processed by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, an anomaly detection algorithm based on machine learning is adopted to identify the potential fault modes of the high-medium pressure valve in real time; according to the potential fault modes, a corresponding maintenance and repair plan is generated based on the decision tree algorithm, and the maintenance and repair plan is feedback to the operator through a mobile application to realize the intelligent monitoring of the high-medium pressure valve. Thus, by using an embedded multi-sensor network, an adaptive data fusion algorithm, and an anomaly detection means based on machine learning, not only can the real-time monitoring and fault warning of the valve be realized, but also it is helpful to optimize the maintenance and repair plan and improve the intelligent level of valve management.

[0106] The embodiment of the present invention also provides a storage medium, in which a computer program is stored. Among them, the computer program is set to execute the steps in any one of the above method embodiments when running.

[0107] Specifically, in this embodiment, the above storage medium may be set to store a computer program for executing the following steps:

[0108] S201, according to the real-time operating state of the high-medium pressure valve and the environmental pressure data, collect data through an embedded multi-sensor network, and comprehensively process the collected data by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics. Among them, the multi-sensor network at least includes a pressure sensor, a temperature sensor, and a vibration sensor;

[0109] S202. Based on the working characteristics of the valve, adopt an anomaly detection algorithm based on machine learning to identify potential fault modes of the high and medium pressure valves in real time;

[0110] S203. According to the potential fault modes, generate corresponding maintenance and repair plans based on the decision tree algorithm, and feedback the maintenance and repair plans to the operator through the mobile application to achieve intelligent monitoring of the high and medium pressure valves.

[0111] It can be seen that according to the real-time operating status and ambient pressure data of the high and medium pressure valves, data is collected through an embedded multi-sensor network, and the collected data is comprehensively processed using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, an anomaly detection algorithm based on machine learning is adopted to identify potential fault modes of the high and medium pressure valves in real time; according to the potential fault modes, corresponding maintenance and repair plans are generated based on the decision tree algorithm, and the maintenance and repair plans are feedback to the operator through the mobile application to achieve intelligent monitoring of the high and medium pressure valves. Thus, it is possible to use the embedded multi-sensor network, the adaptive data fusion algorithm and the anomaly detection means based on machine learning to not only achieve real-time monitoring and fault warning of the valves, but also help to optimize the maintenance and repair plans and improve the intelligent level of valve management.

[0112] An embodiment of the present invention also provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0113] Specifically, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0114] Specifically, in this embodiment, the above processor may be configured to execute the following steps through a computer program:

[0115] S201. According to the real-time operating status and ambient pressure data of the high and medium pressure valves, collect data through an embedded multi-sensor network, and comprehensively process the collected data using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics, wherein the multi-sensor network includes at least a pressure sensor, a temperature sensor and a vibration sensor;

[0116] S202. Based on the valve working characteristics, adopt an anomaly detection algorithm based on machine learning to identify potential fault modes of the high and medium pressure valves in real time;

[0117] S203. Generate a corresponding maintenance and repair plan based on the decision tree algorithm according to the potential failure mode, and feedback the maintenance and repair plan to the operator through the mobile application to achieve intelligent monitoring of high and medium pressure valves.

[0118] It can be seen that according to the real-time operating status and environmental pressure data of high and medium pressure valves, data is collected through an embedded multi-sensor network, and the collected data is comprehensively processed using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics; based on the valve working characteristics, an anomaly detection algorithm based on machine learning is used to identify the potential failure modes of the high and medium pressure valves in real time; according to the potential failure modes, a corresponding maintenance and repair plan is generated based on the decision tree algorithm, and the maintenance and repair plan is feedback to the operator through the mobile application to achieve intelligent monitoring of high and medium pressure valves. Thus, by using an embedded multi-sensor network, an adaptive data fusion algorithm, and an anomaly detection means based on machine learning, not only can real-time monitoring and fault warning of valves be achieved, but also it helps to optimize the maintenance and repair plan and improve the intelligent level of valve management.

[0119] The above has detailed the structure, features, and function effects of the present invention according to the illustrated embodiments. The above is only the preferred embodiment of the present invention, but the present invention is not limited to the scope defined by the drawings. Any changes made according to the concept of the present invention, or equivalent embodiments modified into equivalent changes, still within the spirit covered by the specification and drawings, shall be within the protection scope of the present invention.

Claims

1. An intelligent monitoring method applied to high-medium pressure valves, characterized in that, The method includes: Collect data through an embedded multi-sensor network according to the real-time operating status of the high and medium pressure valves and environmental pressure data, and comprehensively process the collected data using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics. Among them, the multi-sensor network includes at least a pressure sensor, a temperature sensor, and a vibration sensor; for the data stream of each sensor, apply continuous wavelet transform to extract features at different scales, where the features represent the change details and trends of the signal to capture multi-level information of the data; construct a perception game model, regard each sensor as a participant in the game, and each sensor participates in the game according to the contribution degree and reliability of the data it collects, competing for importance in the fusion process. Through iterative calculation, a Nash equilibrium point is reached, at which point the data weight distribution of each sensor reaches the optimal state; input the different scale features from different sensors extracted into a fuzzy logic system, and output a fused feature as the comprehensive valve working characteristic. Among them, the fuzzy rule set is adjusted in real time according to the weights assigned by the perception game model to adapt to different operating states and environments, and perform a fuzzy inference process. Through the matching of the fuzzy set and the fuzzy rule, the different scale features of the sensor are transformed into a comprehensive fused feature, reflecting the valve working state; Based on the valve working characteristics, adopt an anomaly detection algorithm based on machine learning to real-time identify potential fault modes of the high and medium pressure valves; According to the potential fault modes, generate corresponding maintenance and repair plans based on the decision tree algorithm, and feedback the maintenance and repair plans to the operators through a mobile application to achieve intelligent monitoring of the high and medium pressure valves.

2. The method according to claim 1, wherein The adopting an anomaly detection algorithm based on machine learning to real-time identify potential fault modes of the high and medium pressure valves based on the valve working characteristics includes: Input the valve working characteristics into a deep autoencoder network to learn its low-dimensional representation, and use the output of the bottleneck layer of the autoencoder as a feature embedding. This embedding represents the core features of the data and can distinguish normal and abnormal modes. Among them, the autoencoder captures the potential structural features of the data through the reconstruction error; Based on the feature embedding, use the isolation forest algorithm for anomaly detection. The isolation forest constructs decision trees by randomly selecting features and split points to identify abnormal data points different from the normal mode; Use a causal inference network to analyze the causal relationship between the anomaly detection results of the isolation forest algorithm and the operating conditions of the high and medium pressure valves, and identify potential fault modes. Among them, the causal inference network reveals the driving factors behind the abnormal phenomena through a causal graph model; 3. The method according to claim 2, wherein The generating corresponding maintenance and repair plans based on the decision tree algorithm according to the potential fault modes includes: Use one-hot encoding to represent the potential fault modes as binary vectors, and combine with the numerical features of the collected data to form a complete feature vector as the input of the decision tree model; Starting from the root node of the decision tree, traverse the tree structure layer by layer. For each node, use the feature division condition of the node to compare the feature vectors. Among them, if a certain feature value in the feature vector meets a certain preset condition, continue to traverse downward along the branch that meets the preset condition. If a certain feature value does not meet a certain preset condition, continue to traverse downward along the other branch until reaching the leaf node; After path tracing, finally reach a leaf node. This leaf node represents specific maintenance and repair suggestions, and a complete maintenance and repair plan is generated based on the leaf nodes reached by tracing.

4. An intelligent monitoring system applied to high-medium pressure valves, characterized in that, The system includes: An acquisition module, which is used to collect data through an embedded multi-sensor network according to the real-time operating status of high and medium pressure valves and environmental pressure data, and comprehensively process the collected data by using an adaptive data fusion algorithm to obtain comprehensive valve working characteristics. Among them, the multi-sensor network at least includes a pressure sensor, a temperature sensor and a vibration sensor; among them, for the data stream of each sensor, continuous wavelet transform is applied to extract features at different scales. The features represent the change details and trends of the signal to capture multi-level information of the data; construct a perception game model, regard each sensor as a participant in the game, and each sensor participates in the game according to the contribution degree and reliability of the data it collects, competing for importance in the fusion process. Through iterative calculation, a Nash equilibrium point is reached. At this point, the data weight distribution of each sensor reaches the optimal state; input the different scale features from different sensors extracted into the fuzzy logic system, and output a fusion feature as the comprehensive valve working characteristic. Among them, the fuzzy rule set is adjusted in real time according to the weights assigned by the perception game model to adapt to different operating states and environments, and the fuzzy inference process is executed. Through the matching of the fuzzy set and the fuzzy rule, the different scale features of the sensor are transformed into comprehensive fusion features, reflecting the valve working state; An identification module, which is used to adopt an anomaly detection algorithm based on machine learning to identify potential fault modes of the high and medium pressure valves in real time based on the valve working characteristics; A monitoring module, which is used to generate a corresponding maintenance and repair plan based on the decision tree algorithm according to the potential fault mode, and feedback the maintenance and repair plan to the operator through a mobile application to realize intelligent monitoring of high and medium pressure valves.

5. The system according to claim 4, wherein The identification module is specifically used for: Input the valve working characteristics into a deep autoencoder network to learn its low-dimensional representation, and use the output of the bottleneck layer of the autoencoder as a feature embedding. This embedding represents the core features of the data and can distinguish normal and abnormal modes. Among them, the autoencoder captures the potential structural features of the data through the reconstruction error; Based on the feature embedding, use the isolation forest algorithm for anomaly detection. The isolation forest constructs a decision tree by randomly selecting features and split points to identify abnormal data points different from the normal mode; Using a causal inference network, analyze the causal relationship between the anomaly detection results of the isolation forest algorithm and the operating conditions of high and medium pressure valves, and identify potential failure modes. Among them, the causal inference network reveals the driving factors behind abnormal phenomena through a causal graph model.

6. The system according to claim 5, characterized in that, The monitoring module is specifically configured to: Use one-hot encoding to represent the potential failure mode as a binary vector, and combine it with the numerical features of the collected data to form a complete feature vector as the input of the decision tree model; Starting from the root node of the decision tree, traverse the tree structure layer by layer. For each node, use the feature division condition of the node to compare the feature vector. Among them, if a certain feature value in the feature vector meets a certain preset condition, continue to traverse downward along the branch that meets the preset condition. If a certain feature value does not meet a certain preset condition, then continue to traverse downward along the other branch until reaching the leaf node; After path tracing, finally reach a leaf node, which represents specific maintenance and repair suggestions, and generate a complete maintenance and repair plan according to the leaf nodes reached by tracing.

7. A storage medium, characterized in that, A computer program is stored in the storage medium, wherein the computer program is set to execute the method according to any one of claims 1-3 when running.

8. An electronic device, comprising a memory and a processor, characterized in that A computer program is stored in the memory, and the processor is set to run the computer program to execute the method according to any one of claims 1-3.

Images