Measurement point monitoring method and measurement point management system based on correlation analysis

Abstract

The invention relates to the field of power plant measuring point monitoring, in particular to a measuring point monitoring method and a measuring point management system based on correlation analysis, and the method comprises the steps: carrying out the standardized coding processing of mass measuring point data; establishing a communication environment of a plurality of measuring point data sources and a real-time data platform; acquiring measuring point data, and transmitting the measuring point data to a real-time data platform; establishing a measuring point data dynamic analysis and verification model, and processing and analyzing the measuring point data; and intelligent matching query of measuring point data is realized. According to the invention, the monitoring effect of power plant managers on abnormal measuring points can be improved.

Description

Technical Field

[0001] The present invention relates to the field of power plant measuring point monitoring, and in particular to a measuring point monitoring method and a measuring point management system based on correlation analysis. Background Art

[0002] With the rapid development and progress of the power industry, power plants are moving towards informatization and intelligence. In order to understand and control the operating conditions of power plant units in real time, many measuring points composed of sensors are set up between power plant units. The contents measured by these measuring points include temperature, pressure, flow, valve opening and other interrelated physical quantities. The abnormal data in these measuring point data have a great impact on the accuracy of power plant unit parameter prediction, performance monitoring and fault diagnosis. Therefore, in the actual operation and parameter prediction of the unit, unit performance monitoring and fault diagnosis modeling process, detecting and monitoring the measuring point data, and processing the abnormal data in these measuring point data are of great significance to ensure the stable and normal operation of the power plant unit and improve the economy of the power plant operation.

[0003] Most power plant measurement point data is revised by power plant managers on relevant data platforms, or corrected by checking with DCS data, or sorting out equipment indicators at power plant business application measurement points. The above methods require measurement point managers to spend a lot of energy to check the measurement points, and cannot guarantee the timely update of measurement points and the validity of data. Summary of the invention

[0004] In order to improve the management ability of power plant managers on abnormal measurement point data, the present application provides a measurement point monitoring method and a measurement point management system based on correlation analysis.

[0005] In a first aspect, the present application provides a measurement point monitoring method based on correlation analysis, comprising the following steps:

[0006] S1. Standardized coding processing of massive measurement point data;

[0007] S2, build a communication environment between multiple measurement point data sources and real-time data platform;

[0008] S3, obtain the measurement point data, and transmit the measurement point data to the real-time data platform;

[0009] S4. Establish a dynamic analysis and verification model for measuring point data, and process and analyze the measuring point data;

[0010] S5, realize intelligent matching query of measurement point data;

[0011] The method of establishing a dynamic analysis and verification model for measuring point data and processing and analyzing the measuring point data includes the following steps: determining the logical relationship of the dynamic association between the operating parameters during the operation of the unit; establishing a dynamic analysis and verification model for measuring point data based on the multi-dimensional data anomaly detection method of XGBoots to detect abnormal values ​​of measuring points in real time; performing intelligent proofreading on the abnormal values ​​of the detected measuring points and displaying the results to the user;

[0012] The XGBoots-based multi-dimensional data anomaly detection method establishes a dynamic analysis and verification model for measuring point data, and detects abnormal values ​​of measuring points in real time, including the following steps: data collection and cleaning; standardization of the cleaned measuring point data to unify the dimensions of measuring point data of different dimensions; feature extraction and dimensionality reduction; dynamic analysis and verification model training for measuring point data; online anomaly detection, and anomaly monitoring and management.

[0013] The beneficial effects are:

[0014] The construction of the research and application project of automatic identification and accuracy management system of measurement point anomalies based on correlation analysis can optimize the communication between the real-time data platform and the data sources of each measurement point, improve the transmission speed and quality of massive data in the information system, and provide underlying data support with high reliability, availability and high performance for various production analysis platforms.

[0015] By establishing a standardized system for measuring point information, it is convenient to conduct intelligent matching queries, facilitate queries on the correlation between various data and information, and improve the tolerance and recognition rate of erroneous information. Quality management of measuring point data has significantly improved data integrity, data validity, data timeliness, data authenticity, etc., and data coordination technology can effectively reduce data errors. By monitoring and reminding equipment measuring points of changes, and then automatically searching for associations, it is easy and convenient to deal with changes in various measuring points, improve the efficiency of measuring point management, and reduce the cost of measuring point management. This will promote the comprehensive management and life cycle tracking of equipment measuring point information.

[0016] In the second aspect, the present application provides a measuring point monitoring and management system based on association analysis, which adopts the following technical solution: it includes a measuring point detection module, a processing module and a monitoring management module, and the measuring point detection module and the monitoring management module communicate with the processing module at the same time; the measuring point detection module is used to generate measuring point data of power plant units and equipment, and transmit it to the processing module, and the processing module performs standardized processing on the measuring point data, the processing module transmits the processed measuring point data to the monitoring management module, and the monitoring management module is used to realize the management of the measuring point data.

[0017] The beneficial effects are: by establishing a standardized system for measuring point information, the communication between the real-time data platform and various data sources is optimized, the transmission speed and quality of massive data in the information system are improved, and the underlying data support with high reliability, availability and high performance is provided for various production analysis platforms. The quality management of measuring point data has significantly improved the data integrity, data validity, data timeliness and data authenticity. By monitoring and reminding the changes of equipment measuring points, and then automatically linking and searching, it is easy and convenient to deal with the changes of various measuring points, improve the efficiency of measuring point management, and reduce the cost of measuring point management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easily understood. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-restrictive manner, and the same or corresponding numbers represent the same or corresponding parts.

[0019] Figure 1 It is a flowchart of a measurement point monitoring method based on correlation analysis in the present application;

[0020] Figure 2 It is a flowchart of standardized coding processing of massive measurement point data in a measurement point monitoring method based on correlation analysis in the present application;

[0021] Figure 3 It is a flowchart of building a communication environment between multiple measuring point data sources and a real-time data platform in a measuring point monitoring method based on correlation analysis in the present application;

[0022] Figure 4 It is a flowchart of obtaining measuring point data in a measuring point monitoring method based on correlation analysis in the present application, and transmitting the measuring point data to a real-time data platform;

[0023] Figure 5 It is a flowchart of establishing a dynamic analysis and verification model for measuring point data in a measuring point monitoring method based on correlation analysis of the present application, and processing and analyzing the measuring point data;

[0024] Figure 6 It is a flowchart of realizing intelligent matching query of measuring point data in a measuring point monitoring method based on association analysis of the present application;

[0025] Figure 7 This is a module diagram of a measurement point management system based on association analysis in this application.

[0026] Explanation of the accompanying drawings: 1. Measurement point detection module; 2. Processing module; 3. Monitoring and management module. DETAILED DESCRIPTION

[0027] The present application embodiment discloses a measurement point monitoring method based on correlation analysis, referring to Figure 1-Figure 6 , including the following steps:

[0028] S1. Standardized coding processing of massive measurement point data.

[0029] In order to solve the consistency, accuracy and availability of massive measurement point data and improve the retrieval efficiency of massive measurement point data, it is necessary to perform standardized coding processing on massive measurement point data.

[0030] Different point sensors will generate different point data sources, which may use different formats and encoding methods, resulting in inconsistencies in the transmission and integration of point data. Standardized encoding of massive point data ensures consistency when the point data is transmitted between different systems.

[0031] Measuring point sensors are physical devices. Incorrect data coding or format caused by equipment failure and other reasons may lead to errors in measuring point data analysis, affecting the quality of measuring point data analysis and decision-making. Standardized coding processing of massive measuring point data, formulation of unified data standards, and improvement of the accuracy of measuring point data.

[0032] Non-standardized point data will make it difficult for different real-time data platforms and applications to use the data, reducing the usability of the data. Standardized coding of massive point data ensures that the data follows a unified standard, improves the accessibility and usability of the data, and thus improves the efficiency of retrieval and query of the data.

[0033] In addition, when integrating data between multiple data sources and systems, non-standardized data will lead to additional conversion and cleaning work. Standardized coding of massive measurement point data simplifies the data integration process and improves efficiency.

[0034] In order to realize the standardized coding processing of massive measurement point data, refer to Figure 2 , including the following steps:

[0035] S11. Determine the content of the measurement points.

[0036] Specifically, the measuring point content includes but is not limited to unified measuring point coding data, measuring point related description data, measuring point system data, measuring point equipment data, measuring point location data, etc., so as to achieve a comprehensive description of the operating status of the power plant units.

[0037] S12. Determine the measurement point coding criteria.

[0038] Specifically, the measurement point coding criteria are determined by referring to the CIM modeling guideline requirements defined in the IEC standard in the prior art and the equipment classification coding method of the KKS power plant, and combining the characteristics of the power plant units.

[0039] In addition, the measurement point coding criteria support on-site equipment analysis and expert analysis at the power plant business level, thereby improving the scientific nature of refined equipment management.

[0040] S13. Determine the monitoring objects, and sort out and standardize various types of information on the monitoring objects.

[0041] According to the determined measurement point coding criteria, various types of information of the monitoring object are sorted out and standardized. The monitoring object information involves the unit level, equipment level and measurement point level. Including:

[0042] A. Unit level: Provide thermal power units of 300MW / 600MW / 1000MW and other capacity levels. Newly connected units can be classified into capacity levels according to actual capacity.

[0043] B. Equipment level: Provide conventional equipment for units of each capacity level. The units can modify the provided equipment library according to actual equipment monitoring needs.

[0044] C. Measuring point level: The original measuring points of the unit are coded and described in a standardized manner according to the coding rules to achieve the unified application and display of monitoring data. Among them, the original measuring point data comes from the DCS data; the calculated point measuring point is the processed data of the original measuring point, which is used for analysis and diagnosis, such as average value, normal value, evaluation value, etc.

[0045] S14. Maintenance and Updates

[0046] With the introduction of new equipment and new technologies, the measurement point coding rules are regularly updated to adapt to new needs.

[0047] S2. Build a communication environment between multiple measurement point data sources and real-time data platform

[0048] refer to Figure 3 , realizing the communication process between multiple measurement point data sources and the real-time data platform, involves demand analysis, system design and technology selection.

[0049] S21. Clarify the measurement point data information.

[0050] The measurement point data information includes but is not limited to the data source data type and format, data frequency, and data volume estimation.

[0051] S22. Design system architecture.

[0052] The overall architecture of the design system includes but is not limited to data collection, processing, storage and display.

[0053] S23. Communication protocol selection.

[0054] In the embodiment of the present application, the CoAP protocol is selected to realize the transmission of the measurement point data source. Similarly, the MQTT or AMQP protocol can also be selected to adapt to different message transmission scenarios.

[0055] S24 measurement point data flow design.

[0056] Design the flow path of data in the system, including data collection, transmission, processing and storage.

[0057] S25 technology selection.

[0058] Select appropriate technical tools and platforms for message queues, databases, stream processing engines, and front-end display tools. In the embodiment of the present application, Apache Kafka can be selected as the message queue; InfluxDB can be selected as the database; Apache Flink can be selected as the stream processing engine; and Grafana can be selected as the front-end display tool. Similarly, RabbitMQ can be selected as the message queue; TimescaleDB can be selected as the database; Spark Streaming can be selected as the stream processing engine; and Kibana can be selected as the front-end display tool.

[0059] S26. Configure the message queue.

[0060] Configure message queue services to ensure that data can be transmitted efficiently.

[0061] S57. Develop stream processing logic.

[0062] Clean, transform, and aggregate data.

[0063] S28. Database integration.

[0064] Store the processed data in the database

[0065] S3. Obtain the measurement point data and transmit the measurement point data to the real-time data platform.

[0066] Using the measurement point coding criteria, develop and configure data collectors to collect data from measurement point data sources.

[0067] S4. Establish a dynamic analysis and verification model for measuring point data, and process and analyze the measuring point data.

[0068] The dynamic analysis and verification model of the measuring point data is used to mine and utilize the large amount of measuring point data generated during the operation of the unit. Combined with the characteristics of the measuring point data, the characteristic values ​​of the measuring point data are mined and analyzed, abnormal measuring point data is found, and the abnormal measuring point data is corrected and displayed to the management personnel. Figure 4 .

[0069] S41. Determine the logical relationship of the dynamic association between the operating parameters during unit operation.

[0070] There is a physical correlation between the operating parameters of power plant units and equipment. In the normal production process, the operating parameters can also maintain a logical correlation with each other during the real-time dynamic change. It is precisely because of the correlation between these operating parameters that abnormal parameters can be discovered in time by establishing a reliable dynamic analysis verification model.

[0071] S42. Based on the multi-dimensional data anomaly detection method of XGBoots, a dynamic analysis and verification model of measuring point data is established to detect abnormal values ​​of measuring points in real time.

[0072] S421. Data collection and cleaning.

[0073] Real-time measurement point data is obtained from the data collectors of power plant units and equipment, and then the original multidimensional data is cleaned to remove missing values ​​and outliers.

[0074] S422. The cleaned measuring point data are standardized to unify the dimensions of measuring point data of different dimensions.

[0075] S423. Feature extraction and dimensionality reduction.

[0076] The measured point data after standardization are subjected to feature extraction and dimensionality reduction. The correlation of the original measured point data is calculated, and based on the principal component analysis method, the features of the original measured point data are linearly changed to map the measured point data from multi-dimensional space to low-dimensional space, thereby reducing the dimensionality of the measured point data.

[0077] S424. Dynamic analysis of measurement point data to verify model training.

[0078] The XGBoost method is used to train the reduced-dimensional data and establish a prediction model for equipment abnormalities. Figure 5 .

[0079] S425. Abnormal online detection.

[0080] The measured point data to be predicted is input into the trained measured point data dynamic analysis verification model to obtain the model prediction value. The model prediction value and the actual measured value of the data collector are subtracted to obtain the absolute value. If the absolute value exceeds the given threshold, it is determined that an abnormality has occurred.

[0081] For new measuring points or changes or disappearances of existing measuring points, the real-time data platform will automatically remind and search for related items and give corresponding analysis conclusions.

[0082] According to the similarity principle, the limit change characteristics and curve characteristics of the changed measuring points are analyzed for the corresponding origin to find out the measuring points after the abnormal measuring points are changed.

[0083] S426. Abnormal monitoring management.

[0084] The real-time data platform regularly monitors the associated measuring points of the redundant PI servers in the plant and the energy-saving platform PI servers that obtain plant-side data, promptly detects measuring point changes or measuring points with abnormal data, and automatically performs abnormal monitoring and management.

[0085] Continuously monitor changes in measuring points, including new measuring points or changes and disappearances of existing measuring points. Based on the online streaming cloud computing framework, the real-time data platform can continuously and comprehensively scan the measuring points, and compare the characteristic information of the previous measuring points to draw conclusions about the changes in the measuring points in a timely manner and give reminders.

[0086] S43. Intelligently calibrate abnormal values ​​of detection points and display the results to users.

[0087] For abnormal measuring points, the real-time data platform automatically finds the start time of the measuring point abnormality, or the measuring point updated by the factory after the user enters the time, and compares it based on the point limit, change rate, curve characteristics, etc., and selects the measuring point most similar to the abnormal measuring point and pushes it to the user to update the measuring point origin.

[0088] S5. Realize intelligent matching query of measurement point data.

[0089] Intelligent matching query of standardized information of measurement points based on machine learning and natural semantic association analysis technology involves multiple steps including data preprocessing, feature extraction, model training, and similarity calculation. Figure 6 .

[0090] S51. Preprocess the measurement point data.

[0091] Preprocessing includes data cleaning, word segmentation, and part-of-speech tagging; data cleaning is used to remove useless symbols, stop words, etc.; word segmentation is used to split sentences into meaningful vocabulary units; and part-of-speech tagging is used to assign appropriate parts of speech to each word.

[0092] S52. Extract features from the measurement point data.

[0093] In the embodiment of the present application, the text is converted into a vector form using methods such as TF-IDF, and the word vector is obtained using the Word2Vec model to obtain a better semantic representation. Similarly, the word vector can also be obtained using the BERT model.

[0094] S53. Build a knowledge graph.

[0095] A knowledge graph is constructed from text through entity recognition (NER) and relation extraction (RE) techniques.

[0096] Entity recognition is the process of identifying entities (such as temperature, humidity, etc.) in text.

[0097] Relation extraction is the process of identifying relationships between entities (such as "is located in", "belongs to").

[0098] S54. Measurement point data recognition model training.

[0099] The model is trained using a neural network machine learning algorithm. During the training process, the model learns how to identify standardized measurement point information from the feature vector. Similarly, the model can also be trained using machine learning algorithms such as SVM and random forest.

[0100] S55. Intelligent matching query.

[0101] The user enters a query, and the real-time data platform performs the same preprocessing and feature extraction steps.

[0102] Calculate the similarity between the user query and the entities in the knowledge graph, usually using methods such as cosine similarity. Based on the similarity score, return the most matching standardized measurement point information.

[0103] According to the data and information of the measuring point coding rules, value range, equipment, name, etc., the measuring points are intelligently analyzed and matched. The real-time data platform combines basic vocabulary-level analysis such as part-of-speech recognition, special nouns, synonyms, etc., with advanced intelligent analysis applications such as syntactic structure, topic model, knowledge graph, vector analysis, etc., and finally obtains the matching target of key information and text semantic understanding.

[0104] Implementation principle: Standardize the coding management of measuring points, and use machine learning and natural semantic analysis in artificial intelligence to realize intelligent matching query of standardized information of measuring points. By establishing a standardized system for measuring point information, it is convenient to conduct intelligent matching query, facilitate the query of the correlation between various data and various information, and improve the tolerance and recognition rate of erroneous information.

[0105] This application also discloses a measurement point management system based on correlation analysis, referring to Figure 7 , including a measuring point detection module 1, a processing module 2 and a monitoring and management module 3. Among them, the measuring point detection module 1 and the monitoring and management module 3 communicate with the processing module 2 at the same time; the measuring point detection module 1 is used to generate the measuring point data of the power plant units and equipment, and transmit it to the processing module 2, and the processing module 2 performs standardized processing on the measuring point data, and the processing module transmits the processed data to the monitoring and management module 3, so that the management personnel can realize the operation of the entire management system.

[0106] refer to Figure 7,The processing module 2 includes a public data dictionary submodule, an indicator management submodule, and a measurement point data analysis model submodule.

[0107] Among them, the public data dictionary submodule is used to query the commonly used professional definitions and related functional descriptions of this measuring point management system, and implement enumeration type descriptions on the function page; managers can add, modify, and perform other functions on the common types of this measuring point management system.

[0108] The indicator management submodule is used to establish a unified measurement point naming specification for power plant units and equipment. Through standardized measurement point naming, the measurement point data can be displayed with the system diagram, and the measurement point usage of the business system can be matched, further reducing the workload of measurement point maintenance.

[0109] The measuring point data analysis model submodule is used to analyze and process the measuring point data transmitted from the measuring point detection module 1 to determine normal measuring point data and abnormal measuring point data.

[0110] Specifically, refer to Figure 7 The measurement point data analysis model submodule monitors the characteristics of the measurement point itself and verifies the quality of the measurement point data according to the changes in the relevant characteristics of the measurement point. Among them, the characteristics of the measurement point itself include the limit value, change rate, change frequency and measurement point correlation of the measurement point.

[0111] refer to Figure 7 The monitoring management module 3 includes an interface management submodule, an interface monitoring submodule, a measuring point information management submodule, a measuring point abnormality monitoring submodule, a measuring point alarm criterion maintenance submodule, an alarm statistics query submodule and a display submodule.

[0112] The measuring point real-time monitoring submodule communicates with the measuring point abnormality monitoring submodule, is used to monitor the measuring point in real time, and transmit the information to the measuring point abnormality monitoring submodule.

[0113] The interface management submodule is used by managers to manage the power plant interface and realize the unified management of communication parameters between the real-time data platform and the various data source systems of the power plant. The configuration parameters mainly include: communication protocol, communication mode, port address, etc. The interface function includes the power plant communication protocol, and can be managed and called through a graphical interface.

[0114] The interface monitoring submodule is used to realize real-time monitoring of interfaces, interface history status query, interface alarm processing, interface information maintenance, and interface measurement point maintenance. Managers can monitor the measurement point data transmission interface in real time through a graphical interface, and the page can query the interface operation status and data transmission status in real time.

[0115] The interface history status query specifically allows managers to query the historical status of the interface over a period of time, the number of interruptions, interruption time statistics, etc. Managers can click on a single piece of information to display all interface alarm and interruption information for the selected time period, including the interruption start time, end time, and alarm type.

[0116] The specific interface alarm processing is that the system pushes the alarm interface information to relevant personnel according to the interface configuration interface alarm notification type. In this way, authorized users can analyze the interface alarm according to the pushed alarm information, fill in the processing measures, and end the alarm when the processing is completed.

[0117] Specifically, the management personnel can add, modify, delete, and other operations on the interface information through the interface configuration page information; the interface information should include: interface name, type, IP, port, data packet file, parsing method, etc. The management personnel can query the interface configuration, query the original measurement point of the interface and the standard point of entering PISERVER.

[0118] Interface measurement point information maintenance: Administrator users can edit the measurement point information. The measurement point information is updated to the data table by the background program. Users can change whether the measurement point is written to PI or whether the measurement point is disabled.

[0119] The measurement point information management submodule is used to manage the basic information and characteristic information of the measurement points, and the management personnel can add, delete, modify, import and export the basic information and characteristic information of the measurement points. Among them, the basic information management of the measurement points includes the management of basic information such as the measurement point name, original measurement point, point source, equipment, measurement point location, measurement point type information, and classifies the relevant attributes of the measurement points for easy retrieval. The characteristic information of the measurement points includes characteristics, characteristic values, characteristic limits, start time, end time, whether there are related measurement points and other information.

[0120] The measuring point abnormality monitoring submodule is used to query the abnormalities occurring at the measuring points during real-time operation in real time, and to perform historical query and related statistics on the alarm results according to the unit, measuring point level, alarm type, etc.

[0121] The measurement point real-time monitoring submodule can display the current alarm measurement point, including the measurement point name, alarm duration, alarm type, etc.

[0122] The administrator of the measurement point alarm criterion maintenance submodule can click on the measurement point criterion to query the measurement point criterion and edit the measurement point criterion.

[0123] Measuring point characteristics: As the properties of measuring points, measuring point characteristics save the characteristics of measuring points and characteristic judgment limits, which can be used directly as judgment criteria. Measuring point judgment criteria can also be combined and logically judged. Users can change and edit the logic of the judgment criteria.

[0124] The alarm statistics query submodule supports querying alarms within the selected time, counting the number of alarms at the measuring point, alarm duration and other information; and can filter and query abnormal measuring points by unit, system and equipment.

[0125] The display submodule is used to display PI server operation information, network transmission rate, to-do information, interface status and measurement point alarm information.

[0126] Among them, PI server operation information includes the measurement point information of the current PI server, such as the total number of measurement points, regularly updated measurement points, and measurement points that have not been updated for a long time; it includes monitoring of snapshots, history, cache, calculation and other services of PISERVER services; it includes performance monitoring of the PI server host, CPU, memory, storage and other information.

[0127] Among them, the network transmission rate is the transmission of the measurement points of the interfaces in the power plant units and equipment systems to the slave PI server, including the total number of transmission measurement points and the real-time transmission rate. When you click [History], you can view the transmission rate history curve.

[0128] Among them, the pending tasks include but are not limited to interface alarm processing, measurement point alarm confirmation, measurement point update, etc. The manager clicks on the pending items to proceed to the corresponding function page.

[0129] The interface information displays the total received data packets of all interfaces, interface snapshots and historical conditions, and related information of each interface, including interface status, number of received measurement points, data packet size, number of written PI measurement points, etc. The administrator can perform real-time monitoring of the corresponding interface by clicking on the interface.

[0130] Among them, the measuring point alarm information is statistically classified according to the measuring point level and alarm classification; the management personnel can click this function to enter the measuring point real-time alarm function page.

[0131] Implementation principle: Detection module 1 is used to generate measurement point data of power plant units and equipment, and transmit it to processing module 2, which standardizes the measurement point data. The processing module transmits the processed data to monitoring management module 3, so that management personnel can operate the entire management system.

[0132] Although this specification has shown and described a plurality of embodiments of the present application, it is obvious to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, modifications and alternatives without departing from the thought and spirit of the present application. It should be understood that in the process of practicing the present application, various alternatives to the embodiments of the present application described herein may be adopted.

[0133] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A point monitoring method based on correlation analysis, characterized in that: Includes steps: S1. Standardized coding processing of massive measurement point data; S2, build a communication environment between multiple measurement point data sources and real-time data platform; S3, obtain the measurement point data, and transmit the measurement point data to the real-time data platform; S4. Establish a dynamic analysis and verification model for measuring point data, and process and analyze the measuring point data; S5, realize intelligent matching query of measurement point data; The method of establishing a dynamic analysis and verification model for measuring point data and processing and analyzing the measuring point data includes the following steps: determining the logical relationship of the dynamic association between the operating parameters during the operation of the unit; establishing a dynamic analysis and verification model for measuring point data based on the multi-dimensional data anomaly detection method of XGBoots to detect abnormal values ​​of measuring points in real time; Intelligently check abnormal values ​​of test points and display the results to users; The XGBoots-based multi-dimensional data anomaly detection method establishes a dynamic analysis and verification model for measuring point data, and detects abnormal values ​​of measuring points in real time, including the following steps: data collection and cleaning; standardization of the cleaned measuring point data to unify the dimensions of measuring point data of different dimensions; feature extraction and dimensionality reduction; dynamic analysis and verification model training for measuring point data; online anomaly detection, and anomaly monitoring and management.

2. The method for monitoring measurement points based on correlation analysis according to claim 1, characterized in that: The standardized coding process for the massive measurement point data includes the steps of: formulating the measurement point content; determining the measurement point coding criteria; determining the monitoring object, and sorting out and standardizing various types of information of the monitoring object.

3. A method for monitoring measuring points based on correlation analysis according to claim 2, characterized in that: The communication environment of building multiple measurement point data sources and real-time data platform includes the steps of: clarifying measurement point data information; designing system architecture; Communication protocol selection; measurement point data flow design; Technology selection; configure message queues; develop stream processing logic; integrate databases; transfer multiple measurement point data sources to the real-time data platform.

4. The method for monitoring measuring points based on correlation analysis according to claim 3 is characterized in that: The method for realizing intelligent matching query of measuring point data comprises the steps of: preprocessing measuring point data; extracting features of measuring point data; constructing a knowledge graph; training a measuring point data recognition model; and intelligent matching query.

5. According to the method for monitoring measuring points based on correlation analysis in claim 1, the abnormal monitoring management is based on an online streaming cloud computing framework, which continuously and comprehensively scans the measuring points, and combines the characteristic information of the measuring points to draw conclusions on the changes of the measuring points and give reminders.

6. According to the method for monitoring measuring points based on correlation analysis in claim 2, the measuring point coding criteria refer to the CIM modeling guidelines defined in the IEC standard and the equipment classification coding method of the KKS power plant.

7. According to the measurement point monitoring method based on association analysis as described in claim 2, the measurement point content includes measurement point coding data, measurement point related description data, measurement point system data, measurement point equipment data and measurement point location data.

8. A measurement point monitoring and management system based on correlation analysis, characterized in that: It comprises a measuring point detection module (1), a processing module (2) and a monitoring management module (3), wherein the measuring point detection module (1) and the monitoring management module (3) communicate with the processing module (2) at the same time; The measuring point detection module (1) is used to generate measuring point data of power plant units and equipment, and transmit the data to the processing module (2). The processing module (2) performs standardization processing on the measuring point data. The processing module (2) transmits the processed measuring point data to the monitoring management module (3). The monitoring management module (3) is used to manage the measuring point data.

Images