Well station instrument fault intelligent diagnosis system and diagnosis method based on big data and artificial intelligence

Abstract

The invention discloses a well station instrument fault intelligent diagnosis system and diagnosis method based on big data and artificial intelligence, and the system comprises a well station instrument data management and acquisition module which collects and stores well station instrument data in real time, and provides a visual data display interface at the same time, so that an operator can know the operation state of a well station in time; the well station instrument data analyzing and processing module is used for analyzing the collected well station instrument data and extracting key index data; according to the well station instrument fault early warning model system, a multi-level early warning model is created, and potential faults of well station instruments can be effectively recognized and predicted; and the well station instrument fault diagnosis and analysis module is used for quickly identifying abnormal conditions in the well station instrument data, making fault early warning and analysis reports and providing the fault early warning and analysis reports to management personnel. According to the intelligent diagnosis system and the diagnosis method, problems possibly occurring in production are analyzed and predicted through an intelligent algorithm and a model, and corresponding measures are taken in advance so as to ensure safety and stability of oil well production.

Description

Technical Field

[0001] The present invention relates to an intelligent diagnosis system for well station instrument faults, and in particular to an intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence. The present invention also relates to an intelligent diagnosis method for well station instrument faults based on big data and artificial intelligence, and belongs to the technical field of smart oil fields. Background Art

[0002] In the oil industry, well station instruments are key equipment for monitoring and controlling oil well production. Since oilfield well site temperature and pressure instruments, dynamometers, RTUs and other production instruments are in the harsh environment of the field for a long time, they are prone to failure, resulting in the inability to accurately and timely feedback production data to the central control platform, which brings difficulties to production management.

[0003] Existing instrument fault diagnosis technology mainly relies on manual inspection and processing, which cannot meet the needs of smart oilfield construction. In terms of data collection and processing, existing well station instruments rely on manual methods to analyze and diagnose data, which is time-consuming and labor-intensive, making it difficult to monitor a large number of well stations in real time. In addition, traditional data analysis methods lack sufficient intelligence and cannot process massive amounts of data and conduct in-depth analysis. In terms of fault diagnosis, the existing instrument monitoring system was not designed to fully consider future data growth and analysis needs, making it difficult for the system to cope with the growing amount of data and the need to analyze complex problems.

[0004] In the context of the rapid development of informatization and intelligence, building advanced instrument fault diagnosis and early warning models is of great significance to reducing on-site labor and improving production efficiency. The present invention provides an intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence, aiming to realize the intelligence and systematization of oil field development and improve the work efficiency of workers in the oil-related fields. Summary of the invention

[0005] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the above problems and / or the problems existing in the prior art, the present invention is proposed.

[0007] The primary purpose of the present invention is to overcome the problems existing in the prior art and provide an intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence. Through efficient data management, intelligent algorithms and model analysis, it can monitor the status of well station instruments in real time, predict and diagnose potential faults in advance, thereby improving the efficiency and safety of oil well production and facilitating the construction of smart oil fields.

[0008] In order to solve the above technical problems, the intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence of the present invention includes:

[0009] Well station instrument data management and acquisition module collects and stores well station instrument data in real time, and provides an intuitive data display interface;

[0010] Well station instrument data analysis and processing module analyzes the collected well station instrument data and extracts key indicator data;

[0011] Well station instrument failure early warning model system, creating a multi-level early warning model, which can effectively identify and predict potential failures of well station instruments;

[0012] The well station instrument fault diagnosis and analysis module can quickly identify abnormal conditions in the well station instrument data, and produce fault warning and analysis reports to provide to management personnel.

[0013] Furthermore, the intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence also includes:

[0014] The system monitoring and optimization module monitors and evaluates the overall performance of the well station instrument fault warning model. The key evaluation indicators include model accuracy, response speed and stability. The parameters and algorithms in the model are corrected after regular evaluation to ensure the stability and accuracy of the early warning system.

[0015] Furthermore, the well station instrument data analysis and processing module uses association rule mining and data preprocessing technology to analyze the collected well station instrument data and extract indicator data that can be used by the well station instrument fault warning model system.

[0016] Furthermore, the well station instrument fault warning model system combines the K-means clustering analysis method in traditional machine learning with the time series feature extraction capability of the CNN-LSTM network in deep learning to create a multi-level warning model.

[0017] Another object of the present invention is to overcome the problems existing in the prior art and provide an intelligent diagnosis method for well station instrument faults based on big data and artificial intelligence, which can completely and comprehensively manage and analyze the well station instrument data collected by the production unit, analyze and predict possible problems in production through intelligent algorithms and models, and take corresponding measures in advance to ensure the safety and stability of oil well production, promote the intelligence and systematization of oil field development, and help improve the work efficiency of personnel in oil-related fields.

[0018] In order to solve the above technical problems, the intelligent diagnosis method of well station instrument fault based on big data and artificial intelligence of the present invention comprises the following steps in sequence:

[0019] S1, real-time collection and storage of well station instrument data;

[0020] S2. Conduct in-depth analysis on the collected well station instrument data and extract key indicator data;

[0021] S3, integrating traditional machine learning methods and deep learning technology to create a multi-level early warning model;

[0022] S4. Analyze the acquired well station instrument data in real time, quickly determine the abnormal situation, and prepare fault warning and analysis reports to provide to management personnel;

[0023] S5. Continuously monitor and optimize the early warning model.

[0024] Furthermore, step S1 specifically includes the following sub-steps:

[0025] S1.1, Instrument identification and tracking: Unique electronic tag encoding is performed on all instrument equipment, and corresponding electronic files are established to record the usage and maintenance record information of the instrument; the electronic tag traceability technology is used to quickly identify the instrument, and the combination of RTU feedback positioning and RFID scanning positioning is used to quickly locate the instrument information;

[0026] S1.2. Establish a complete data management system to manage instrument data.

[0027] Furthermore, step S1.2 specifically includes the following sub-steps:

[0028] S1.2.1, Data collection and storage: by connecting with instrument equipment, collect instrument data in real time and store it in a unified database;

[0029] S1.2.2, Data query and display: Provide a user-friendly query interface to facilitate users to query and retrieve data according to their needs; at the same time, visualize the data;

[0030] S1.2.3, Data security and authority control: Use the DES algorithm to encrypt data and take backup and recovery measures. At the same time, use the authority control mechanism to manage the authority of different users.

[0031] Furthermore, the specific steps of extracting key indicator data in step S2 are as follows:

[0032] S2.1. Clean the basic data of the instrument and remove obvious errors and mismatched data;

[0033] S2.2. Extract features related to the problem from the basic data through feature extraction algorithms, statistical methods or machine learning techniques;

[0034] S2.3. According to the pre-set standard data format, the classified basic data are standardized to obtain consistent and comparable indicator data.

[0035] Furthermore, the creation of a multi-level early warning model in step S3 includes the following sub-steps in sequence:

[0036] S3.1. For the obtained indicator data, apply the K-means algorithm to divide the instrument equipment data into clusters with similar characteristics, and find outliers and regular patterns of instrument operation;

[0037] S3.2. Use the CNN network to extract the high-dimensional feature vectors of the above instrument-related data, construct the obtained high-dimensional feature vectors into a time series, and use them as the input data of the LSTM network;

[0038] S3.3, apply the LSTM algorithm to the modeling of time series data, and predict the data in the future by training the LSTM model, so as to realize the early warning of abnormal instrument conditions;

[0039] S3.4. Build a CNN-LSTM network.

[0040] Furthermore, the K-means algorithm in step S3.1 includes the following steps in sequence:

[0041] S3.1.1. Select k initialized instrument data samples as initial cluster centers a=a1,a2,…a k ;

[0042] S3.1.2. For each sample x in the instrument data set i Calculate its distance to k cluster centers and divide it into the class corresponding to the cluster center with the smallest distance;

[0043] S3.1.3 For each category a j , recalculate its cluster center That is, the centroid of all samples belonging to this class;

[0044] S3.1.4. Repeat steps S3.1.2 and S3.1.3 above until the number of iterations is reached or the algorithm terminates when the minimum error changes.

[0045] Furthermore, the CNN-LSTM network in step S3.4 includes the following layers:

[0046] S3.4.1, the first layer is the input layer;

[0047] S3.4.2, the second layer is the convolutional neural network layer;

[0048] S3.4.3, the third layer is a multi-layer LSTM;

[0049] S3.4.4, the fourth layer is the attention layer;

[0050] S3.4.5. The fifth layer is the output layer.

[0051] Furthermore, in step S3.4.1, the format of the input data is specified, the batch size is set to 1 by default, the number of time steps is t, and the feature dimension is n. Then a sample is represented as a real number sequence matrix Rt×n, and xi is recorded as the vector representation of the i-th time step data in Rt×n;

[0052] In step S3.4.2, the sample data enters the CNN layer and performs convolution, pooling and dimensionality reduction operations in sequence;

[0053] In step S3.4.3, the output of the previous LSTM layer is the input of the next layer, and it is passed down layer by layer. The output of the last LSTM hidden layer enters the attention layer for further processing.

[0054] S3.4.4, the LSTM hidden layer output vector is used as the input of the attention layer, trained through the fully connected layer, and then the output of the fully connected layer is normalized using the softmax function to obtain the distribution weight of each hidden layer vector. The weight size indicates the importance of the hidden state of each time step to the prediction result;

[0055] S3.4.5, the output layer specifies the prediction time step 0 t , and finally output 0 t The prediction results of the step.

[0056] Further, in step S3.4.2,

[0057] One-dimensional convolution is adopted, and the convolution kernel is convolved only in a single time domain direction; the number of convolution kernels is r, the size is set to k, and X1:i+k-1 is the real number matrix from the i-th time step to the i+k-1-th time step in Rt×n, and the sliding step size is 1;

[0058] The weight matrix W1 is a k×n real number matrix; a feature extraction is performed on the sequence vector of each k time step to obtain a feature 0 i , the calculation formula is as follows:

[0059]

[0060] f is a nonlinear activation function, b1∈R is a bias; when a convolution kernel extracts a sample sequence data, a feature map of (t-k+1)×1 shape is obtained, and the calculation formula is as follows:

[0061] 0=[01,02,…,0 t-k+1 ] T

[0062] CNN has a total of r convolution kernels, so r feature maps will be obtained in the end; after convolution, the maximum pooling operation is performed, the pooling size is 2, the sliding step is 2, and r feature maps of the shape of [(t-k+1) / 2]×1 are obtained. The calculation formula is as follows:

[0063] 0=max{01,0 1+1}(1=1,3,5,…,tk)

[0064] These r feature maps are the features extracted by the CNN layer. They are reduced in dimension into a real vector of length r*(t-k+1) / 2, which stores the spatial connection between different eigenvalues ​​in the sample data and is then input into the LSTM layer for further processing.

[0065] Furthermore, in step S3.4.4, the weight training process is as follows:

[0066] S i =tanh(WH1+b1)

[0067] α i =softmax(S i )

[0068] Then use the trained weights to calculate the weighted average sum of the hidden layer output vectors. The calculation results are as follows:

[0069]

[0070] Among them, H i is the output of the last LSTM hidden layer, S i is the score for each hidden layer output, α i is the weight coefficient, C1 is the result of weighted summation, and softmax is the activation function.

[0071] Furthermore, step S4 includes the following sub-steps in sequence:

[0072] S4.1. Identify abnormal situations: Use the early warning model to analyze the acquired well station instrument data in real time, receive the early warning results from the well station instrument fault early warning model system, and then parse and analyze the early warning results to determine whether there is a real abnormal situation;

[0073] S4.2. Determine the warning level: Determine the corresponding warning level for each warning situation based on the output and real-time data of the well station instrument failure warning model system;

[0074] S4.3. Formulate response measures: formulate corresponding response measures according to the warning level;

[0075] S4.4. Send warning notification: After determining the warning level and response measures, send a warning notification to relevant personnel.

[0076] Furthermore, the specific scheme of step S5 for continuously monitoring and optimizing the early warning model includes:

[0077] S5.1. Monitor the performance of the well station instrument fault warning model system: By monitoring the various indicators and operating status of the well station instrument fault warning model system, understand the performance of the system, including the accuracy, recall rate, and false alarm rate indicators of the well station instrument fault warning model system, as well as the real-time response time and resource utilization of the system, timely discover potential problems and performance bottlenecks of the system, and take corresponding measures to optimize them.

[0078] Furthermore, the specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes:

[0079] S5.2. Evaluate the effectiveness of the well station instrument failure warning model system: Regularly evaluate the effectiveness and accuracy of the well station instrument failure warning model system. By comparing with actual abnormal conditions and events, analyze the consistency between the predicted results of the well station instrument failure warning model system and the actual situation, and evaluate the accuracy and reliability of the model. If deviations or errors are found in the well station instrument failure warning model system, make corresponding adjustments and improvements based on the evaluation results.

[0080] Furthermore, the specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes:

[0081] S5.3. Optimize the parameters and algorithms of the well station instrument failure warning model system: Based on the evaluation and monitoring results of the warning model effect, optimize the parameters and algorithms of the model as needed; improve the prediction ability and stability of the model by adjusting the model parameter settings, selecting a more appropriate algorithm or introducing new feature engineering methods. This module can also use historical data and feedback information, and use machine learning technology to automatically optimize and update the well station instrument failure warning model system.

[0082] Furthermore, the specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes:

[0083] S5.4. Introduce advanced models and algorithms: Introduce deep learning models and reinforcement learning algorithms; at the same time, combine the experience and knowledge of domain experts to integrate and fuse models.

[0084] Compared with the prior art, the present invention has achieved the following beneficial effects: 1. Improved the production efficiency of oilfield well stations: This system effectively manages instrument equipment and data by tracking and locating instrument equipment, as well as real-time collection, storage, query and display of instrument data, greatly improving the production efficiency of oilfield well stations and ensuring the continuity and stability of the production process. Due to the huge number of instruments in the domestic oil industry and the huge workload of manual inspections, the application of this instrument fault intelligent diagnosis system and its diagnosis method can greatly improve production efficiency and reduce maintenance costs. It has a very broad application prospect and provides an excellent technical approach to accelerate the construction of smart oil fields.

[0085] 2. Improved the production and economic benefits of oilfield well stations: This system uses big data analysis technology to process and optimize the monitoring data obtained from oilfield well station instruments, which can help optimize the production process; by better understanding and utilizing production data, oil fields can manage production in a refined manner and improve production and benefits.

[0086] 3. Real-time monitoring and predictive maintenance of oilfield well stations: This system can detect abnormal conditions before equipment failures by real-time monitoring of the status of various instrumentation equipment at the well station, which helps to implement predictive maintenance, carry out repairs and replacements in a timely manner, and reduce the impact of equipment failures on production.

[0087] 4. Reduced oilfield operation and maintenance costs: By achieving fine management of instrumentation and data, the maintenance and management costs of oilfields are reduced, thereby reducing the operating costs of oilfields. Through predictive maintenance, this system can reduce the frequency of emergency repairs and equipment replacement, thereby reducing maintenance costs; accurate fault prediction and effective maintenance strategies help extend equipment life and reduce the maintenance burden of enterprises.

[0088] 5. Enhanced the safety of oilfield well stations: Through real-time collection and analysis of well station instrument data, combined with traditional machine learning and deep learning technologies, a multi-level instrument early warning model system has been built, which can predict and identify abnormal situations in real time, effectively identify and predict well site instrument failures, and take corresponding response measures in a timely manner, effectively avoiding potential safety risks and ensuring the production safety of oilfield well stations.

[0089] 6. Improved oil field reliability and reduced production downtime: Through predictive maintenance and timely fault warning, the system can reduce production downtime caused by instrument failure. This is crucial for the oil field production industry because any production interruption may lead to economic losses.

[0090] 7. Continuous monitoring and optimization functions: By introducing advanced models and algorithms, optimizing model parameters and algorithms, the performance and stability of the early warning model system are continuously improved. Continuous performance improvement ensures the reliability of oilfield well station production and further reduces failure rate and downtime. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor. The accompanying drawings are only provided for reference and explanation, and are not used to limit the present invention.

[0092] in:

[0093] Figure 1 This is a diagnostic flow chart of the intelligent diagnostic system for well station instrument faults based on big data and artificial intelligence of the present invention;

[0094] Figure 2 This is a workflow diagram of the well station instrument data management and acquisition module in the present invention;

[0095] Figure 3 This is a workflow diagram of the well station instrument data analysis and processing module in the present invention;

[0096] Figure 4 This is a working diagram of the well station instrument failure early warning model system in the present invention;

[0097] Figure 5 This is a working diagram of the fault diagnosis and analysis module of the well station instrument in the present invention;

[0098] Figure 6 It is a working flow chart of the system monitoring and optimization module in the present invention;

[0099] Figure 7 This is an interface diagram when the system of the present invention diagnoses "dynamometer load sensor failure" in Yong 52-4 well;

[0100] Figure 8 This is an interface diagram of the system of the present invention when diagnosing the fault of "dynamometer - instrument recent data volume reduction" in well Lian 6-3;

[0101] Fig. 9 This is an interface diagram when the system of the present invention diagnoses the fault of "dynamometer-instrument day and night data difference" in Lian 6-2 and other wells;

[0102] Fig.10 This is an interface diagram when the system of the present invention diagnoses the fault of "oil pressure gauge - instrument day and night data difference" in Cema 24 well. DETAILED DESCRIPTION

[0103] In the following description of the present invention, the directions or positional relationships indicated by terms such as "up", "down", "front", "back", "left", "right", "inside" and "outside" are based on the directions or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description, and do not mean that the device must have a specific direction.

[0104] In order to make the technical means, creative features, objectives and effects of the present invention easier to understand, the present invention is further described below in conjunction with specific diagrams. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them.

[0105] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0106] like Figures 1 to 6 As shown, the intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence of the present invention includes a well station instrument data management and acquisition module, a well station instrument data analysis and processing module, a well station instrument fault early warning model system, a well station instrument fault diagnosis and analysis module and a system monitoring and optimization module, aiming to improve the data management and fault diagnosis capabilities of oilfield well station instrument equipment and solve the deficiencies of the prior art in data processing, fault prediction and system optimization.

[0107] The well station instrument data management and acquisition module is responsible for real-time collection and storage of well station instrument data, ensuring data accuracy and integrity through advanced data processing technology, while providing an intuitive data display interface to facilitate operators to understand the well station operation status in a timely manner. It realizes the data collection, storage, analysis and display of various instrument equipment at oilfield well stations, ensuring real-time monitoring of instrument equipment status and reasonable management of data, and providing reliable data support and decision-making basis for oilfield well station production operations.

[0108] The well station instrument data analysis and processing module conducts in-depth analysis on the collected well station instrument data, and uses advanced algorithms such as association rule mining and data preprocessing technology to extract key indicator data, that is, to obtain indicator data that can be used by the final early warning model, providing a reliable data basis for subsequent fault early warnings.

[0109] The well station instrument fault warning model system integrates traditional machine learning methods and deep learning technology to create a multi-level warning model. This model combines the K-means clustering analysis method in traditional machine learning with the time series feature extraction capability of the CNN-LSTM network in deep learning, and can effectively identify and predict potential faults of well station instruments.

[0110] The well station instrument fault diagnosis and analysis module uses the well station instrument fault warning model to perform real-time analysis on various well station instrument data in oil field production obtained by the well station instrument data analysis and processing module, quickly identify abnormal conditions in the well station instrument data, and prepare fault warning and analysis reports to provide to management personnel.

[0111] The system monitoring and optimization module is a key module for continuous monitoring and optimization of the well station instrument fault warning model system. It is responsible for monitoring and evaluating the overall performance of the well station instrument fault warning model system, including key indicators such as model accuracy, response speed and stability. The parameters and algorithms in the model are corrected after regular evaluation to ensure continuous optimization and upgrading of the system to ensure the stability and accuracy of the warning system. At the same time, the module can also gradually introduce more advanced models and algorithms according to actual application conditions to adapt to the ever-changing production environment and data requirements, continuously improve the performance and effect of the well station instrument fault warning model, and further improve the intelligence level of the entire system.

[0112] Through the above series of technical solutions, the present invention can realize efficient management and intelligent fault diagnosis of oilfield well station instrument equipment, improve production operation safety and efficiency, and provide strong technical support for the sustainable development of oilfields.

[0113] like Figure 2 As shown in the figure, the well station instrument data management and acquisition module ensures data accuracy and integrity through data processing technology, including:

[0114] S1.1, Instrument identification and tracking: To ensure the safety of production and operation of oilfield well stations, all instruments and equipment are coded with unique electronic tags, and corresponding electronic files are established to record the usage and maintenance record information of the instruments. Through electronic tag traceability technology, instruments can be quickly identified, and according to the on-site process flow and instrument equipment installation of the oilfield well station, the combination of RTU return positioning and RFID scanning positioning can be used to quickly locate instrument information and ensure the production and operation of the oilfield well station.

[0115] S1.2, Instrument data management: In order to better manage instrument data, a complete data management system is established, which has the following functions:

[0116] S1.2.1, Data collection and storage: Through the connection with the instrument equipment, the instrument data is collected in real time and stored in a unified database. The data collection can include the real-time data, historical data, alarm and fault information of the instrument. The collected data is reasonably stored and archived for subsequent data analysis and query.

[0117] S1.2.2, Data query and display: Provide a user-friendly query interface to facilitate users to query and retrieve data according to their needs. Flexible queries can be performed according to time range, instrument type, location and other conditions to quickly obtain the required data. At the same time, data is visualized through charts, reports and other methods, so that users can intuitively understand the changes and trends of data.

[0118] S1.2.3, Data security and authority control: Ensure the security and credibility of instrument data, use the DES algorithm to encrypt data and take backup and recovery measures to prevent data loss or tampering. At the same time, through the authority control mechanism, manage the authority of different users to ensure that data can only be accessed and operated by authorized personnel, and protect the confidentiality and integrity of data.

[0119] like Figure 3 As shown in the figure, the specific steps of extracting key indicator data by the well station instrument data analysis and processing module are as follows:

[0120] S2.1. Clean the basic data of the instrument and remove obvious errors and mismatched data. The cleaning process includes but is not limited to the detection and deletion of data outliers, filling or deleting missing data, etc., to ensure the accuracy and consistency of the data.

[0121] S2.2. Perform feature extraction on the preprocessed basic data to extract useful key information. Through feature extraction algorithms, statistical methods or machine learning techniques, extract features related to the problem from the basic data, including key attributes, trend patterns, abnormal behaviors, etc. The extracted features are used for subsequent classification, analysis, modeling and other tasks.

[0122] S2.3. According to the pre-set standard data format, the classified basic data is standardized to obtain consistent and comparable indicator data. Standardization includes operations such as data scaling, normalization, and discretization, which make different features comparable and facilitate subsequent data analysis, model building, and decision support applications. At the same time, standardization also helps to reduce data dimensions, reduce redundant information, and improve the efficiency of data processing and analysis.

[0123] like Figure 4 As shown, the well station instrument failure early warning model system creates a multi-level early warning model, which includes the following steps in sequence:

[0124] S3.1. For the obtained indicator data, the K-means algorithm is applied to divide the instrument equipment data into clusters with similar characteristics, and find abnormal points and regular patterns of instrument operation. These abnormal points may be instruments with problems, while regular patterns can be used to identify instruments that are operating normally.

[0125] The algorithm steps of K-means are:

[0126] S3.1.1. Select k initialized instrument data samples as initial cluster centers a=a1,a2,…a k ;

[0127] S3.1.2. For each sample x in the instrument data set i Calculate its distance to k cluster centers and divide it into the class corresponding to the cluster center with the smallest distance;

[0128] S3.1.3 For each category a j , recalculate its cluster center That is, the centroid of all samples belonging to this class;

[0129] S3.1.4. Repeat steps S3.1.2 and S3.1.3 above until the number of iterations is reached or the algorithm terminates when the minimum error changes.

[0130] S3.2. Use CNN network to extract high-dimensional feature vectors of the above instrument-related data: By performing convolution and pooling operations on text data, CNN network extracts feature information of different scales and angles to form high-dimensional feature vectors. The obtained high-dimensional feature vectors are constructed into time series and used as input data for LSTM network. When constructing time series, it is necessary to consider the setting of time step and sliding window size, as well as maintaining the continuity and consistency of data. LSTM network uses the high-dimensional feature vectors extracted by the previous CNN network to model and predict instrument data.

[0131] S3.3. The LSTM algorithm is used to effectively capture the long-term dependencies in time series data, and has good modeling and prediction capabilities. In the instrument warning model system, the LSTM algorithm is applied to the modeling of time series data. By training the LSTM model, the data in the future can be predicted, thereby realizing the warning of abnormal instrument conditions.

[0132] S3.4. Construct a CNN-LSTM network. The CNN-LSTM network includes the following layers:

[0133] S3.4.1. The first layer is the input layer. The format of the input data (batch size, time steps, feature dimension) is specified. The batch size is set to 1 by default, the time step is t, and the feature dimension is n. Then a sample is represented by a real number sequence matrix Rt×n, and xi is the vector representation of the i-th time step data in Rt×n.

[0134] S3.4.2, the second layer is the convolutional neural network layer, namely the CNN layer. The CNN layer can extract the spatial connection between different eigenvalues ​​in the data, thereby making up for the shortcomings of the LSTM network that cannot capture the spatial components of the data. At the same time, the features it extracts are still temporal. The sample data enters the CNN layer and performs convolution, pooling and node expansion (dimensionality reduction) operations in sequence. For sequence data, this model adopts one-dimensional convolution, and the convolution kernel only performs convolution in a single time domain direction. The number of convolution kernels is r, the size is set to k, and X1:i+k-1 is the real number matrix from the i-th time step to the i+k-1-th time step in Rt×n, and the sliding step size is 1.

[0135] The weight matrix W1 is a k×n real number matrix. A feature extraction is performed on the sequence vector of each k time step to obtain a feature 0 i , the calculation formula is as follows:

[0136]

[0137] f is a nonlinear activation function, and b1∈R is a bias. When a convolution kernel extracts a sample sequence data, a feature map of (t-k+1)×1 shape is obtained, and the calculation formula is as follows:

[0138] 0=[01,02,…,0 i-k+1 ] T

[0139] CNN has a total of r convolution kernels, so it will eventually get r feature maps. After convolution, the maximum pooling operation is performed, the pooling size is 2, the sliding step is 2, and r feature maps of the shape [(t-k+1) / 2]×1 are obtained. The calculation formula is as follows:

[0140] 0=max{01,0 1+1}(1=1,3,5,…,tk)

[0141] These r feature maps are the features extracted by the CNN layer. They are reduced in dimension into a real vector of length r*(t-k+1) / 2, which stores the spatial connection between different eigenvalues ​​in the sample data and is then input into the LSTM layer for further processing.

[0142] S3.4.3, the third layer is a multi-layer LSTM. LSTM has a memory function and can extract the time series change information of instrument fault nonlinear data. It introduces input gate, forget gate, output gate, and also adds candidate state, cell state and hidden state. Cell state stores long-term memory, which can alleviate gradient disappearance, and hidden state stores short-term memory. This model uses multi-layer LSTM. The output of the previous LSTM layer is the input of the next layer, which is passed down layer by layer. The output of the last LSTM hidden layer enters the attention layer for further processing.

[0143] S3.4.4, the fourth layer is the attention layer. Attention can enhance the role of important time steps in LSTM, thereby further reducing the model prediction error. Attention is essentially to find the weighted average sum of the output vectors of the last layer of LSTM. The output vector of the LSTM hidden layer is used as the input of the attention layer, trained through the fully connected layer, and then the output of the fully connected layer is normalized using the softmax function to obtain the assigned weight of each hidden layer vector. The weight size indicates the importance of the hidden state of each time step to the prediction result. The weight training process is as follows:

[0144] S i =tanh(WH1+b1)

[0145] α i =softmax(S i )

[0146] Then use the trained weights to calculate the weighted average sum of the hidden layer output vectors. The calculation results are as follows:

[0147]

[0148] Among them, H i is the output of the last LSTM hidden layer, S i is the score for each hidden layer output, α i is the weight coefficient, C1 is the result of weighted summation, and softmax is the activation function.

[0149] S3.4.5, the fifth layer is the output layer. This layer specifies the prediction time step 0 t , and finally output 0 t The prediction results of the step.

[0150] Combining the characteristics of K-means and CNN-LSTM, a comprehensive instrument failure warning model system can be constructed. First, the K-means algorithm is used to cluster the instrument data to find outliers and regular patterns. Then, for each cluster, the CNN-LSTM algorithm is used to model and predict time series data to obtain future trends and abnormalities. Finally, by comprehensively analyzing the results of K-means and CNN-LSTM, early warning judgment and decision-making of instrument failures are carried out.

[0151] The early warning model system can monitor the instrument data of oilfield well stations in real time, detect abnormal situations in time and predict future trends, which helps managers to respond quickly and take corresponding measures to reduce risks and improve the production efficiency and safety of oilfields. At the same time, in the actual operation of the model system, the accuracy and reliability of the early warning model can be improved by continuously optimizing and improving the algorithm.

[0152] like Figure 5 As shown in the figure, the specific working steps of the well station instrument fault diagnosis and analysis module are as follows:

[0153] S4.1. Identify abnormal situations: Use the early warning model to analyze the acquired well station instrument data in real time and receive the early warning results from the well station instrument fault early warning model system. These early warning results may include abnormal status of the instrument, trend changes, or indicator data exceeding the set threshold. The early warning decision module will parse and analyze these early warning results to determine whether there is a real abnormal situation.

[0154] S4.2. Determine the warning level: Determine the corresponding warning level for each warning situation based on the output and real-time data of the well station instrument fault warning model system. The warning level can be evaluated and classified according to factors such as the severity of the anomaly, the scope of impact, and the duration. Generally, the warning level can be divided into different levels, such as low, medium, and high, to facilitate subsequent response and processing.

[0155] S4.3. Develop response measures: Develop corresponding response measures according to the warning level. These measures may include adjusting instrument parameters, starting backup equipment, performing maintenance and repair, and adding personnel. The warning decision module can select the most appropriate response measures to deal with abnormal situations based on historical data and experience knowledge, combined with the current situation and warning level, and ensure the timeliness and effectiveness of the actions taken.

[0156] S4.4. Send warning notifications: After determining the warning level and response measures, send warning notifications to relevant personnel. These notifications can be sent via email, SMS, mobile applications, etc. to ensure that relevant personnel can receive them in time and take necessary actions. The notification content includes the warning level, description of the abnormal situation, recommended response measures, and contact information, etc., so that relevant personnel can conduct further processing and communication.

[0157] Through the operation of the early warning decision module, oilfield managers can be notified in time when an early warning occurs and take appropriate measures to deal with abnormal situations, thereby reducing risks and losses. At the same time, the early warning decision module can also record early warning information and response processes for subsequent analysis and optimization. This real-time early warning and decision-making capability will help improve the production efficiency, safety and reliability of oilfields.

[0158] like Figure 6 As shown in the figure, the specific process of the system monitoring and optimization module to continuously monitor and optimize the well station instrument fault warning model includes:

[0159] S5.1. Monitor the performance of the well station instrument fault warning model system: By monitoring the various indicators and operating status of the well station instrument fault warning model system, the performance of the system can be understood. For example, the accuracy, recall rate, false alarm rate and other indicators of the well station instrument fault warning model system, as well as the real-time response time, resource utilization and other operating status of the system can be monitored to timely discover potential problems and performance bottlenecks of the system, and take corresponding measures to optimize them.

[0160] S5.2. Evaluate the effectiveness of the well station instrument failure warning model system: The system monitoring and optimization module will regularly evaluate the effectiveness and accuracy of the well station instrument failure warning model system. By comparing with actual abnormal conditions and events, analyze the consistency between the predicted results of the well station instrument failure warning model system and the actual situation, and evaluate the accuracy and reliability of the model. If deviations or errors are found in the well station instrument failure warning model system, make corresponding adjustments and improvements based on the evaluation results.

[0161] S5.3. Optimize the parameters and algorithms of the well station instrument failure warning model system: Based on the evaluation and monitoring results of the warning model effect, optimize the parameters and algorithms of the model as needed. Improve the prediction ability and stability of the model by adjusting the model parameter settings, selecting a more appropriate algorithm, or introducing new feature engineering methods. This module can also use historical data and feedback information, and use machine learning technology to automatically optimize and update the well station instrument failure warning model system.

[0162] S5.4. Introducing advanced models and algorithms: In order to continuously improve the performance and effectiveness of the early warning system, the system monitoring and optimization module can gradually introduce more advanced models and algorithms. For example, deep learning models, reinforcement learning algorithms, etc. can be introduced to better explore the potential laws and characteristics of the data. At the same time, combined with the experience and knowledge of experts in the field, model integration and fusion are carried out to further improve the prediction accuracy and stability of the early warning system. Through the operation of the system monitoring and optimization module, the early warning model system can detect and correct problems in a timely manner to ensure the stability and accuracy of the system. The process of optimizing model parameters and algorithms will continuously improve the performance of the early warning model, enabling it to better adapt to actual application scenarios. At the same time, the introduction of advanced models and algorithms will further promote the development of the early warning system, improve its prediction accuracy and ability to cope with complex situations.

[0163] like Figure 7 As shown, the diagnostic system of the present invention automatically diagnosed that "dynamometer load sensor failure" occurred in Yong 52-4 well.

[0164] like Figure 8 As shown, the diagnostic system of the present invention automatically diagnosed the interface diagram of "recent data volume reduction of dynamometer instruments" in wells such as Lian 6-3, Lian 30-3, and Cehua 24-3; subsequent inspections revealed that some instruments had antenna failures and some instruments had battery power shortages, which caused a decrease in instrument data transmission volume, and all of them were dealt with in a timely manner.

[0165] like Fig. 9 As shown, the diagnostic system of the present invention automatically diagnoses the "day and night data difference of the dynamometer" of wells such as Lian 6-2, Lian 46-5, and Hua 26-6. The reason for the day and night data difference is that the power of the solar cell is not enough for a whole night, resulting in the data being able to be uploaded normally during the day when there is light; in the second half of the night, the data cannot be uploaded due to the power exhaustion of the solar cell.

[0166] like Fig.10 As shown, the diagnostic system of the present invention automatically diagnosed the "oil pressure gauge - instrument day and night data difference" fault of Cema 24 well. The reason is that the power of the solar cell used by the oil pressure gauge is not enough for a whole night, resulting in the data being able to be uploaded normally when there is light during the day; in the second half of the night, the data cannot be uploaded due to the power exhaustion of the solar cell.

[0167] The above is only a preferred embodiment of the present invention, which shows and describes the basic principles, main features and advantages of the present invention, but does not limit the scope of patent protection of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. In addition to the above embodiments, the present invention may have other implementation modes without departing from the spirit and scope of the present invention. The present invention may also have various changes and improvements, and all technical solutions formed by equivalent substitution or equivalent transformation fall within the scope of protection required by the present invention. The scope of protection required by the present invention is defined by the attached claims and their equivalents. The technical features not described in the present invention can be realized by or using existing technologies, which will not be repeated here.

Claims

1. An intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence, characterized in that: include: Well station instrument data management and acquisition module collects and stores well station instrument data in real time, and provides an intuitive data display interface; Well station instrument data analysis and processing module analyzes the collected well station instrument data and extracts key indicator data; Well station instrument failure early warning model system, creating a multi-level early warning model, which can effectively identify and predict potential failures of well station instruments; The well station instrument fault diagnosis and analysis module can quickly identify abnormal conditions in the well station instrument data, and produce fault warning and analysis reports to provide to management personnel.

2. The intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence according to claim 1 is characterized in that: Also includes: The system monitoring and optimization module monitors and evaluates the overall performance of the well station instrument fault warning model. The key evaluation indicators include model accuracy, response speed and stability. The parameters and algorithms in the model are corrected after regular evaluation to ensure the stability and accuracy of the early warning system.

3. The intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence according to claim 1 is characterized by: The well station instrument data analysis and processing module uses association rule mining and data preprocessing technology to analyze the collected well station instrument data and extract indicator data that can be used by the well station instrument fault early warning model system.

4. The intelligent diagnosis system for well station instrument faults based on big data and artificial intelligence according to claim 1 is characterized by: The well station instrument failure early warning model system combines the K-means clustering analysis method in traditional machine learning with the time series feature extraction capability of the CNN-LSTM network in deep learning to create a multi-level early warning model.

5. An intelligent diagnosis method for well station instrument faults based on big data and artificial intelligence, characterized in that: The steps are as follows: S1, real-time collection and storage of well station instrument data; S2. Conduct in-depth analysis on the collected well station instrument data and extract key indicator data; S3, integrating traditional machine learning methods and deep learning technology to create a multi-level early warning model; S4. Analyze the acquired well station instrument data in real time, quickly determine the abnormal situation, and prepare fault warning and analysis reports to provide to management personnel; S5. Continuously monitor and optimize the early warning model.

6. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 5 is characterized in that: Step S1 specifically includes the following sub-steps: S1.1, Instrument identification and tracking: Unique electronic tag encoding is performed on all instrument equipment, and corresponding electronic files are established to record the usage and maintenance record information of the instrument; the electronic tag traceability technology is used to quickly identify the instrument, and the combination of RTU feedback positioning and RFID scanning positioning is used to quickly locate the instrument information; S1.

2. Establish a complete data management system to manage instrument data.

7. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 6 is characterized in that: Step S1.2 specifically includes the following sub-steps: S1.2.1, Data collection and storage: by connecting with instrument equipment, collect instrument data in real time and store it in a unified database; S1.2.2, Data query and display: Provide a user-friendly query interface to facilitate users to query and retrieve data according to their needs; at the same time, visualize the data; S1.2.3, Data security and authority control: Use the DES algorithm to encrypt data and take backup and recovery measures. At the same time, use the authority control mechanism to manage the authority of different users.

8. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 5 is characterized in that: The specific steps of extracting key indicator data in step S2 are as follows: S2.

1. Clean the basic data of the instrument and remove obvious errors and mismatched data; S2.

2. Extract features related to the problem from the basic data through feature extraction algorithms, statistical methods or machine learning techniques; S2.

3. According to the pre-set standard data format, the classified basic data are standardized to obtain consistent and comparable indicator data.

9. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 5 is characterized in that: The creation of a multi-level early warning model in step S3 includes the following sub-steps in sequence: S3.

1. For the obtained indicator data, apply the K-means algorithm to divide the instrument equipment data into clusters with similar characteristics, and find outliers and regular patterns of instrument operation; S3.

2. Use the CNN network to extract the high-dimensional feature vectors of the above instrument-related data, construct the obtained high-dimensional feature vectors into a time series, and use them as the input data of the LSTM network; S3.3, apply the LSTM algorithm to the modeling of time series data, and predict the data in the future by training the LSTM model, so as to realize the early warning of abnormal instrument conditions; S3.

4. Build a CNN-LSTM network.

10. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 9 is characterized in that: The K-means algorithm in step S3.1 includes the following steps in sequence: S3.1.

1. Select k initialized instrument data samples as initial cluster centers a=a1,a2,…a j ; S3.1.

2. For each sample x in the instrument data set i Calculate its distance to k cluster centers and divide it into the class corresponding to the cluster center with the smallest distance; S3.1.3 For each category a j , recalculate its cluster center That is, the centroid of all samples belonging to this class; S3.1.

4. Repeat steps S3.1.2 and S3.1.3 above until the number of iterations is reached or the algorithm terminates when the minimum error changes.

11. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 9 is characterized in that: The CNN-LSTM network in step S3.4 includes the following layers: S3.4.1, the first layer is the input layer; S3.4.2, the second layer is the convolutional neural network layer; S3.4.3, the third layer is a multi-layer LSTM; S3.4.4, the fourth layer is the attention layer; S3.4.

5. The fifth layer is the output layer.

12. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11 is characterized in that: In step S3.4.1, the format of the input data is specified, the batch size is set to 1 by default, the number of time steps is t, and the feature dimension is n. Then a sample is represented by a real sequence matrix Rt×n, and xi is the vector representation of the i-th time step data in Rt×n; In step S3.4.2, the sample data enters the CNN layer and performs convolution, pooling and dimensionality reduction operations in sequence; In step S3.4.3, the output of the previous LSTM layer is the input of the next layer, and it is passed down layer by layer. The output of the last LSTM hidden layer enters the attention layer for further processing. S3.4.4, the LSTM hidden layer output vector is used as the input of the attention layer, trained through the fully connected layer, and then the output of the fully connected layer is normalized using the softmax function to obtain the distribution weight of each hidden layer vector. The weight size indicates the importance of the hidden state of each time step to the prediction result; S3.4.5, the output layer specifies the prediction time step 0 t , and finally output 0 t The prediction results of the step.

13. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 12 is characterized in that: In step S3.4.2, a one-dimensional convolution is adopted, and the convolution kernel is convolved only in a single time domain direction; the number of convolution kernels is r, the size is set to k, and X 1: i+k-1 is the real number matrix from the i-th time step to the i+k-1-th time step in Rt×n, and the sliding step length is 1; The weight matrix W1 is a k×n real number matrix; a feature extraction is performed on the sequence vector of each k time step to obtain a feature 0 i , the calculation formula is as follows: f is a nonlinear activation function, b1∈R is a bias; when a convolution kernel extracts a sample sequence data, a feature map of (t-k+1)×1 shape is obtained, and the calculation formula is as follows: 0＝[01,02,…,0 t-k+1 ] T CNN has a total of r convolution kernels, so r feature maps will be obtained in the end; after convolution, the maximum pooling operation is performed, the pooling size is 2, the sliding step is 2, and r feature maps of the shape of [(t-k+1) / 2]×1 are obtained. The calculation formula is as follows: 0=max{01,0 1+1 }(1=1,3,5,…,t-k) These r feature maps are the features extracted by the CNN layer. They are reduced in dimension into a real vector of length r*(t-k+1) / 2, which stores the spatial connection between different eigenvalues ​​in the sample data and is then input into the LSTM layer for further processing.

14. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 12 is characterized in that: In step S3.4.4, the weight training process is as follows: S i =tanh(WH1+b1) α i =softmax(S i ) Then use the trained weights to calculate the weighted average sum of the hidden layer output vectors. The calculation results are as follows: Among them, H i is the output of the last LSTM hidden layer, S i is the score for each hidden layer output, α i is the weight coefficient, C1 is the result of weighted summation, and softmax is the activation function.

15. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11, characterized in that: Step S4 includes the following sub-steps in sequence: S4.

1. Identify abnormal situations: Use the early warning model to analyze the acquired well station instrument data in real time, receive the early warning results from the well station instrument fault early warning model system, and then parse and analyze the early warning results to determine whether there is a real abnormal situation; S4.

2. Determine the warning level: Determine the corresponding warning level for each warning situation based on the output and real-time data of the well station instrument failure warning model system; S4.

3. Formulate response measures: formulate corresponding response measures according to the warning level; S4.

4. Send warning notification: After determining the warning level and response measures, send a warning notification to relevant personnel.

16. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11, characterized in that: The specific scheme of step S5 for continuously monitoring and optimizing the early warning model includes: S5.

1. Monitor the performance of the well station instrument fault warning model system: By monitoring the various indicators and operating status of the well station instrument fault warning model system, understand the performance of the system, including the accuracy, recall rate, and false alarm rate indicators of the well station instrument fault warning model system, as well as the real-time response time and resource utilization of the system, timely discover potential problems and performance bottlenecks of the system, and take corresponding measures to optimize them.

17. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11 is characterized in that: The specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes: S5.

2. Evaluate the effectiveness of the well station instrument failure warning model system: Regularly evaluate the effectiveness and accuracy of the well station instrument failure warning model system. By comparing with actual abnormal conditions and events, analyze the consistency between the predicted results of the well station instrument failure warning model system and the actual situation, and evaluate the accuracy and reliability of the model. If deviations or errors are found in the well station instrument failure warning model system, make corresponding adjustments and improvements based on the evaluation results.

18. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11, characterized in that: The specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes: S5.

3. Optimize the parameters and algorithms of the well station instrument failure warning model system: Based on the evaluation and monitoring results of the warning model effect, optimize the parameters and algorithms of the model as needed; improve the prediction ability and stability of the model by adjusting the model parameter settings, selecting a more appropriate algorithm or introducing new feature engineering methods. This module can also use historical data and feedback information, and use machine learning technology to automatically optimize and update the well station instrument failure warning model system.

19. The intelligent diagnosis method for well station instrument fault based on big data and artificial intelligence according to claim 11, characterized in that: The specific scheme of step S5 for continuously monitoring and optimizing the early warning model also includes: S5.

4. Introduce advanced models and algorithms: Introduce deep learning models and reinforcement learning algorithms; at the same time, combine the experience and knowledge of domain experts to integrate and fuse models.

Images