Thermal power plant auxiliary machine equipment health state monitoring method and system based on polynomial regression

Abstract

The invention discloses a thermal power plant auxiliary machine equipment health state monitoring method and system based on polynomial regression, and relates to the technical field of signal data integration technology and equipment state monitoring, and the method comprises the steps: carrying out the data collection of auxiliary machine equipment based on an Internet of Things platform, mining equipment core characteristic parameters through employing a multiple linear regression algorithm, and carrying out the monitoring of the health state of the auxiliary machine equipment; constructing a typical operation characteristic parameter set; the collected data are preprocessed, an auxiliary machine equipment health state monitoring AI model is constructed based on a polynomial regression algorithm, and the health degree is evaluated; real-time data are input into an AI model for prediction, an AI + mechanism early warning model is built through a logical operation assembly and mechanism modeling, and state judgment and graded early warning are achieved. According to the method, the abnormal state of the auxiliary machine equipment can be timely and accurately detected and dealt with, the safety and stability of the power production process are improved, the engineering practicability is high, the response is faster, the implementation cost is remarkably reduced, and the labor intensity and the working pressure are relieved.

Description

Technical Field

[0001] The present invention relates to the fields of signal data integration technology and equipment status monitoring, and in particular to a method and system for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression. Background Art

[0002] Since the early days of distributed control systems (DCS) in thermal power plants, coordinated control of units has replaced manual operations. Operators use DCS monitoring screens to monitor the operating parameters of auxiliary equipment in real time, make operational adjustments, analyze data, and handle incidents. However, equipment monitoring using DCS is overly dependent on human intervention and lacks in-depth analysis of historical equipment data.

[0003] In the context of digital transformation in the power industry, thanks to the advancement of sensor technology, computer technology and network technology, the widespread application of intelligent algorithms and big data analysis in thermal power plants has made the condition monitoring of auxiliary equipment in coal-fired power plants gradually move towards intelligence.

[0004] On the one hand, the commonly used equipment condition monitoring methods mentioned above utilize complex algorithms to only identify equipment faults offline, failing to achieve real-time early warning and early detection. On the other hand, existing methods only unilaterally predict equipment operating parameters and fail to consider equipment energy consumption from an economic perspective. Therefore, it is of great practical significance to fully integrate equipment operation and economic characteristics, accurately mine historical equipment data features based on a multivariate linear regression algorithm, improve the implementation efficiency of equipment condition monitoring solutions, and thus achieve real-time and efficient health status monitoring of auxiliary equipment in coal-fired power plants. Summary of the Invention

[0005] In view of the above-mentioned problems, the present invention is proposed.

[0006] Therefore, the technical problems solved by the present invention are: the existing auxiliary equipment health status monitoring technology lacks in-depth mining and analysis of the equipment's historical data characteristics, the monitoring method uses complex algorithms and cannot achieve early monitoring and early warning, as well as how to accurately mine the equipment's historical data characteristics and improve the efficiency of the equipment status monitoring solution.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression, comprising collecting data on auxiliary equipment based on an Internet of Things platform, mining the core characteristic parameters of the equipment using a multivariate linear regression algorithm, and constructing a typical operating characteristic parameter set; preprocessing the collected data, constructing an AI model for monitoring the health status of the auxiliary equipment based on a polynomial regression algorithm, and evaluating the health status; inputting real-time data into the AI ​​model for prediction, and building an AI+mechanism early warning model using logical operation components and mechanism modeling to achieve status judgment and graded early warning.

[0008] As a preferred solution of the method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression described in the present invention, the data acquisition includes using a DCS data interface and a SIS system data interface as the data interface, integrating multi-source heterogeneous data of a real-time database, and using an Internet platform and a TrendDB real-time library to realize the acquisition of historical data and real-time data of auxiliary equipment.

[0009] As a preferred solution of the health status monitoring method of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the core characteristic parameters include selecting equipment unit consumption as a key status parameter from an economic perspective and using a relevant analytical algorithm to eliminate honorary measurement points from a safe operation perspective to mine typical operation characteristic parameters.

[0010] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the preprocessing includes, based on the mined typical operating characteristic parameters, performing equipment historical data quality judgment at the edge layer through histogram analysis, data fluctuation judgment, and unreasonable value range judgment based on the Internet platform, and using component operators to achieve data cleaning and improve the quality of training data.

[0011] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the construction of an AI model for monitoring the health status of auxiliary equipment includes mining the correlation between core monitoring parameters and representative parameters of equipment operating conditions through a multivariate linear regression algorithm, fitting the relationship under the historical operating median of the parameters by using quantile sampling, inputting parameter intervals based on the historical data of characteristic parameters collected by the equipment, outputting the equipment core parameter health interval data set according to the operator component, realizing the construction of the AI ​​model for monitoring the health status of auxiliary equipment, and evaluating the health of the constructed AI model for monitoring the health status of auxiliary equipment.

[0012] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the operator components include: a time series definition operator, a database data acquisition operator, a histogram analysis operator, a value range cleaning operator, a data column processing operator, a linear regression model operator, and a model training operator are used as components required in the model construction process; the time series definition operator gives the initial definition of parameters to facilitate the training and calculation of the AI ​​model for monitoring the health status of auxiliary equipment; the data acquisition operator temporarily stores the data in the time series database into the operator by adding the relevant parameter measurement point names through the data acquisition protocol; the histogram analysis operator performs histogram analysis on the data and views the data distribution in a graphical manner; the value range cleaning operator clears abnormal data that exceeds the threshold range by setting a threshold; the linear regression model operator is a basic tool for data training using the multivariate linear regression algorithm; and the model training operator is a tool for executing model training.

[0013] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the health evaluation of the auxiliary equipment health status monitoring AI model includes three evaluation index dimensions: MSE, MAE and R2, and the calculation results of MSE, MAE and R2 are used as evaluation criteria. The lower the values ​​of the MSE and MAE calculation result indicators, the better the performance of the auxiliary equipment health status monitoring AI model, and the higher the value of the R2 calculation result indicator, the better the performance of the auxiliary equipment health status monitoring AI model.

[0014] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the AI+mechanism early warning model includes using the constructed AI model to fit and mine the variable operating condition health operation relationship of the core monitoring parameters of the equipment, predict and output the upper and lower limit thresholds of the parameter operation under real-time conditions, and then perform logical operations and comparisons with the real-time parameter values, and issue prompts and warnings when the real-time values ​​exceed the predicted thresholds to complete the health monitoring of the auxiliary equipment status; the logical operation components include AND, OR, NOT, R / S trigger, M / N, and signal hold.

[0015] As a preferred solution of the method for monitoring the health status of auxiliary equipment of a thermal power plant based on polynomial regression described in the present invention, the auxiliary equipment status includes dividing the equipment operating status according to the operating range of the auxiliary equipment unit consumption; when the coal mill unit consumption is greater than or equal to 4.8kWh / t and less than or equal to 8.2kWh / t, the equipment operating status is a normal operating status; when the coal mill unit consumption is less than 4.8kWh / t, it indicates that the coal mill unit consumption is low, and a low coal mill unit consumption warning operation is executed; when the coal mill unit consumption is greater than 8.2kWh / t, it indicates that the coal mill unit consumption is high, and a high coal mill unit consumption warning operation is executed.

[0016] Another object of the present invention is to provide a health status monitoring system for auxiliary equipment of a thermal power plant based on polynomial regression. By constructing an AI model for health status monitoring of auxiliary equipment and a corresponding AI+ mechanism early warning model, the system can solve the problems of the current equipment status monitoring technology, such as the inability to achieve online real-time monitoring and the lack of consideration of equipment energy consumption economic indicators.

[0017] As a preferred solution of the health status monitoring system of auxiliary equipment of thermal power plant based on polynomial regression described in the present invention, it includes: data acquisition and typical characteristic parameter module, data preprocessing and model construction module, equipment status monitoring and early warning module; the data acquisition and typical characteristic parameter module includes a data acquisition module and a typical characteristic parameter module. The data acquisition module realizes real-time data acquisition through the Internet of Things platform. Data acquisition supports multiple acquisition methods such as DCS data interface, SIS system data interface and real-time database. The Internet platform and TrendDB real-time library are used to realize the collection of historical data and real-time data of auxiliary equipment. The typical characteristic parameter module combines the experience of operation experts to determine the core parameters representing the variable operating condition characteristics of auxiliary equipment, and further fully combines the equipment operation and economic characteristics. From the perspective of economy, some auxiliary equipment selects equipment unit consumption as the key status parameter. From the perspective of safe operation, the correlation analysis algorithm is used to eliminate redundant equipment measurement points with coherence, and typical operation characteristic parameters with high correlation are mined to realize accurate and efficient status monitoring of auxiliary equipment; the data preprocessing and model construction module includes a data preprocessing module and a model construction module The data preprocessing module is based on the Internet platform and uses the edge layer to judge the quality of equipment historical data through multiple methods such as histogram analysis, data fluctuation judgment, and unreasonable value range judgment, and then uses component operators to realize data cleaning and improve the quality of training data. The model construction module includes a multiple linear regression AI model. The multiple linear regression AI model uses a visual modeling tool to quickly build a polynomial regression AI model of 6 major auxiliary equipment by "drag and drop" based on the historical operation data of the past year. According to the floating trend of the historical data of the characteristic parameters of each device, the upper and lower floating ranges of the parameters are set, and finally the healthy range data set of the core parameters of the equipment is output; the equipment status monitoring and early warning module includes an equipment status monitoring module and an equipment early warning module. The equipment status monitoring module compares the healthy range data set output by the constructed multiple linear regression AI model with the real-time data collected by the equipment. The data within the healthy range data set is normal, and the data beyond the healthy range data set is unhealthy. The equipment early warning module uses the conclusion output by the equipment status monitoring module to issue a high warning for the unhealthy equipment if the data exceeds the range, and a low warning for the data below the range.

[0018] A computer device includes a memory and a processor. The memory stores a computer program. The processor executes the computer program to implement a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression.

[0019] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression.

[0020] Beneficial effects of the present invention: The present invention provides a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression. The method collects historical and real-time data of auxiliary equipment through an Internet of Things platform, combines a multivariate linear regression algorithm, and screens typical operating characteristic parameters, thereby improving the safety and stability of the power production process. The polynomial regression algorithm is used to construct an AI model. The algorithm is relatively simple, and the construction efficiency and engineering practicality are improved. The response is faster and the project implementation cost is significantly reduced. The logical operation components and mechanism modeling are used to build an early warning model, establish a unified status monitoring system, improve the work efficiency of the power plant, and reduce the labor intensity and work pressure of the operating personnel. The present invention has achieved better results in terms of the power production process, engineering implementation costs and power plant work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 This is an overall flow chart of a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression, provided as a first embodiment of the present invention.

[0023] Figure 2 A real-time monitoring diagram of the health status of a coal mill in a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression is provided as a second embodiment of the present invention.

[0024] Figure 3 This is an overall flow chart of a health status monitoring system for auxiliary equipment in a thermal power plant based on polynomial regression, provided as a third embodiment of the present invention. DETAILED DESCRIPTION

[0025] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0026] Example 1, with reference to Figure 1 , which is an embodiment of the present invention, provides a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression, comprising:

[0027] S1: Based on the Internet of Things platform, historical and real-time data of auxiliary equipment are collected, and the core feature parameters are screened out using the multivariate linear regression algorithm to construct a typical operation feature parameter set.

[0028] Furthermore, data collection includes using DCS data interface and SIS system data interface to integrate multi-source heterogeneous data of real-time database, and using Internet platform and TrendDB real-time library to realize the collection of historical data and real-time data of auxiliary equipment.

[0029] It should be noted that the core characteristic parameters include selecting equipment unit consumption as the key status parameter from an economic perspective and using relevant analytical algorithms to eliminate honorary measurement points from a safe operation perspective to explore typical operation characteristic parameters.

[0030] It should also be noted that the typical operating characteristic parameters include, depending on the equipment, the typical operating characteristic parameters are different, the typical parameters of the coal mill equipment refer to the coal mill unit consumption, loading oil pressure and primary air pressure, the coal mill unit consumption is the unit mass (usually 1 ton) of qualified coal powder consumed by the power, the unit is kWh / t, the three typical wind pressure parameter values ​​include fan unit consumption and wind pressure, fan unit consumption is the ratio of the power consumed by the primary fan during operation to the mass of the processed coal, the unit is kWh / t, the wind pressure is the inlet wind pressure at the corresponding position.

[0031] S2: Preprocess the collected data, build an AI model for monitoring the health status of auxiliary equipment based on the polynomial regression algorithm, and evaluate the health of the model.

[0032] Furthermore, preprocessing includes, based on the mined typical operating characteristic parameters, using the Internet platform at the edge layer through histogram analysis, data fluctuation judgment, and unreasonable value range judgment to judge the quality of equipment historical data, and using component operators to achieve data cleaning and improve the quality of training data.

[0033] It should be noted that the AI ​​model for monitoring the health status of auxiliary equipment includes mining the correlation between core monitoring parameters and representative parameters of equipment operating conditions through a multivariate linear regression algorithm, fitting the relationship under the historical operating median of the parameters using quantile sampling, inputting parameter intervals based on the historical data of characteristic parameters collected by the equipment, and outputting the equipment core parameter health interval data set according to the operator component to realize the construction of the AI ​​model for monitoring the health status of auxiliary equipment, and evaluating the health of the constructed AI model for monitoring the health status of auxiliary equipment.

[0034] It should also be noted that the operator components include the time series definition operator, database data collection operator, histogram analysis operator, range cleaning operator, data column processing operator, linear regression model operator, and model training operator as the components required in the process of building the AI ​​model for monitoring the health status of auxiliary equipment.

[0035] The time series definition operator provides the initial definition of parameters to facilitate the training and calculation of the AI ​​model for auxiliary equipment health status monitoring. The data acquisition operator adds the names of relevant parameter measurement points through the data acquisition protocol and temporarily stores the data in the time series database into the operator. The histogram analysis operator performs histogram analysis on the data and views the data distribution in a graphical manner. The value range cleaning operator clears abnormal data that exceeds the threshold range by setting the threshold. The linear regression model operator uses the multivariate linear regression algorithm as a basic tool for data training. The model training operator is a tool for executing model training.

[0036] It should also be noted that the health evaluation of the auxiliary equipment health status monitoring AI model includes the three evaluation index dimensions of MSE, MAE and R2, with the calculation results of MSE, MAE and R2 as the evaluation criteria. The lower the value of the MSE and MAE calculation result indicators, the better the performance of the auxiliary equipment health status monitoring AI model, and the higher the value of the R2 calculation result indicator, the better the performance of the auxiliary equipment health status monitoring AI model.

[0037] S3: Through real-time data input into the AI ​​model for prediction, the AI+mechanism warning model is built using logical operation components and mechanism modeling to achieve status judgment and graded warning.

[0038] Furthermore, the AI+mechanism warning model includes using the constructed AI model to fit and mine the variable operating condition health operation relationship of the core monitoring parameters of the equipment, predict and output the upper and lower limit thresholds of the parameter operation under real-time conditions, and then perform logical operations and comparisons with the real-time parameter values, and issue prompts and warnings when the real-time values ​​exceed the predicted thresholds, thereby completing the health monitoring of the auxiliary equipment status.

[0039] Logic operation components include AND, OR, NOT, R / S flip-flop, M / N, and signal hold.

[0040] It should be noted that the auxiliary equipment status includes the classification of equipment operating status according to the operating range of the auxiliary equipment unit consumption; when the coal mill unit consumption is greater than or equal to 4.8kWh / t and less than or equal to 8.2kWh / t, the equipment operating status is the normal operating status; when the coal mill unit consumption is less than 4.8kWh / t, it indicates that the coal mill unit consumption is low, and the coal mill unit consumption low warning operation is executed; when the coal mill unit consumption is greater than 8.2kWh / t, it indicates that the coal mill unit consumption is high, and the coal mill unit consumption high warning operation is executed.

[0041] Example 2, reference Figure 2 , which is an embodiment of the present invention, provides a health status monitoring method for auxiliary equipment in a thermal power plant based on polynomial regression. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0042] First, a plant collected historical data of relevant measurement points for 12 devices, including 6 major auxiliary equipment such as coal mills and primary fans, based on the plant-side DCS system and SIS system data interface. The sampling period was half a year and the sampling interval was 30 seconds. Then, the variable operating condition characteristic parameters and other operating and economic core parameters of the equipment were determined by combining the experience of operation experts and the correlation analysis algorithm. Then, based on the quality analysis and cleaning of the equipment historical data, the polynomial regression algorithm was used to construct an AI model for monitoring the health status of each auxiliary equipment. Finally, based on the industrial Internet platform, an AI+mechanism auxiliary equipment status early warning model was constructed to achieve real-time and efficient monitoring of the health status of each auxiliary equipment. Figure 2 , which can monitor the health status of the coal mill in real time. To evaluate the AI ​​model constructed in this paper, we used the AI ​​model for monitoring the condition of the 1A coal mill in Unit 1 of a power plant as an example. We conducted multiple comparisons with the neural network model using the three statistical evaluation indicators of MSE, MAE, and R², as well as the training time dimension. With reference to Table 1, we recorded and analyzed some experimental data.

[0043] Table 1 Experimental data table

[0044]

[0045]

[0046] As can be seen from the table, compared with the neural network model used for health status monitoring of traditional coal-fired power auxiliary equipment, the AI ​​model constructed by our invention based on multivariate linear regression has a slightly higher model accuracy than the neural network model, and is far superior to the neural network model in terms of model training time. This greatly improves the efficiency in subsequent supplementation of model training data sets, optimization of model parameters and retraining, and can be more effectively used for real-time monitoring of the health status of auxiliary equipment in coal-fired power plants.

[0047] A multivariate linear regression algorithm is used to fit the historical trends of the operating conditions of each auxiliary machine's core monitoring parameters. Further threshold processing is performed to output a historical health data set of each device's monitoring parameters. This data is displayed on the front-end screen as interval bands, with real-time parameter status displayed as red dots. Simultaneously, a parameter early warning model based on the backend AI+ mechanism compares the real-time parameter status with the predicted status, providing early warnings for relevant parameters and rationalizing adjustment recommendations.

[0048] After the auxiliary equipment health monitoring system was put into operation at a thermal power plant, it detected 264 alarms across 14 devices, including six auxiliary equipment, at Unit 1 within a two-month operating cycle. These included multiple equipment alarms, such as "high coal mill unit consumption," "high coal mill primary air pressure," "low air preheater differential pressure," and "high condenser circulating water temperature rise," among other status monitoring parameter warnings. Of these, 230 alarms were correctly and promptly reported to operators, achieving an accuracy rate of over 95% for correct alarms under abnormal condition monitoring. Leveraging precise and efficient predictive models, the auxiliary equipment health monitoring system provides real-time health status monitoring of each auxiliary equipment with a simple, visually appealing display, significantly improving operator efficiency.

[0049] Example 3, reference Figure 3 , which is an embodiment of the present invention, provides a health status monitoring system for auxiliary equipment in a thermal power plant based on polynomial regression, including a data acquisition and typical characteristic parameter module 100, a data preprocessing and model building module 200, and an equipment status monitoring and early warning module 300.

[0050] Among them, S4: data acquisition and typical characteristic parameter module 100 includes data acquisition module 101 and typical characteristic parameter module 102. Data acquisition module 101 realizes real-time data acquisition through the Internet of Things platform. Data acquisition supports multiple acquisition methods such as DCS data interface, SIS system data interface and real-time database. With the help of Internet platform and TrendDB real-time library, the collection of historical data and real-time data of auxiliary equipment is realized. Typical characteristic parameter module 102 combines the experience of operation experts to determine the core parameters representing the variable operating condition characteristics of auxiliary equipment, and further fully combines the equipment operation and economic characteristics. From the perspective of economy, some auxiliary equipment selects equipment unit consumption as the key status parameter. From the perspective of safe operation, the correlation analysis algorithm is used to eliminate redundant equipment measurement points with coherence disclosure, and typical operation characteristic parameters with high correlation are mined to realize accurate and efficient status monitoring of auxiliary equipment.

[0051] Among them, S5: data preprocessing and model building module 200 includes data preprocessing module 201 and model building module 202. Data preprocessing module 201 is based on the Internet platform and uses the edge layer to judge the quality of equipment historical data through multiple methods such as histogram analysis, data fluctuation judgment, and unreasonable value range judgment, and then uses component operators to realize data cleaning and improve the quality of training data. Model building module 202 includes a multivariate linear regression AI model. Based on the historical operation data of the past year, the multivariate linear regression AI model uses a visual modeling tool to quickly build a polynomial regression AI model of 6 major auxiliary equipment by "drag and drop". According to the floating trend of the historical data of the equipment characteristic parameters, the upper and lower floating ranges of the parameters are set, and finally the healthy range data set of the core parameters of the equipment is output;

[0052] Among them, S6: The equipment status monitoring and early warning module 300 includes an equipment status monitoring module 301 and an equipment early warning module 302. The equipment status monitoring module 301 compares the healthy interval data set output by the constructed multivariate linear regression AI model with the real-time data collected by the equipment. The data within the healthy interval data set is in a normal state, and the data beyond the healthy interval data set is in an unhealthy state. The equipment early warning module 302 uses the conclusion output by the equipment status monitoring module 301 to issue a high warning for data above the range and a low warning for data below the range for equipment in an unhealthy state.

[0053] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0054] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0055] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0056] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logical functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc. It should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention and are not intended to be limiting. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced with equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications should be encompassed by the claims of the present invention.

Claims

1. A method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression, characterized in that: include: Based on the data collection of auxiliary equipment on the Internet of Things platform, the core characteristic parameters of the equipment are mined using the multivariate linear regression algorithm to construct a typical operation characteristic parameter set; Preprocess the collected data, build an AI model for monitoring the health status of auxiliary equipment based on the polynomial regression algorithm, and evaluate its health status; By inputting real-time data into the AI ​​model for prediction, and using logical operation components and mechanism modeling to build an AI+mechanism warning model, status judgment and graded warning can be achieved.

2. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 1, characterized in that: The data collection includes: The data interface uses the DCS data interface and SIS system data interface to integrate multi-source heterogeneous data of the real-time database, and uses the Internet platform and TrendDB real-time library to realize the collection of historical data and real-time data of auxiliary equipment.

3. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 1, characterized in that: The core feature parameters include: From the perspective of economy, equipment unit consumption is selected as the key status parameter, and from the perspective of safe operation, relevant analytical algorithms are used to eliminate honorary measurement points and explore typical operation characteristic parameters.

4. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 1, wherein: The pretreatment includes: According to the mined typical operating characteristic parameters, the quality of equipment historical data is judged at the edge layer through histogram analysis, data fluctuation judgment, and unreasonable value range judgment based on the Internet platform. Component operators are used to achieve data cleaning and improve the quality of training data.

5. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 1, characterized in that: The construction of the auxiliary equipment health status monitoring AI model includes: The multivariate linear regression algorithm is used to explore the correlation between the core monitoring parameters and the representative parameters of the equipment operating conditions. The quantile sampling method is used to fit the relationship under the historical operating median of the parameters. According to the historical data of the characteristic parameters collected by the equipment, the parameter interval is input, and the equipment core parameter health interval data set is output according to the operator component to realize the construction of the auxiliary equipment health status monitoring AI model and evaluate the health of the constructed auxiliary equipment health status monitoring AI model.

6. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 5, characterized in that: The operator components include: In the process of building the AI ​​model for monitoring the health status of auxiliary equipment, the time series definition operator, database data collection operator, histogram analysis operator, range cleaning operator, data column processing operator, linear regression model operator, and model training operator are used as the components required in the process of building the AI ​​model for monitoring the health status of auxiliary equipment. The time series definition operator provides the initial definition of parameters to facilitate the training and calculation of the AI ​​model for auxiliary equipment health status monitoring. The data acquisition operator adds the names of relevant parameter measurement points through the data acquisition protocol and temporarily stores the data in the time series database into the operator. The histogram analysis operator performs histogram analysis on the data and views the data distribution in a graphical manner. The value range cleaning operator clears abnormal data that exceeds the threshold range by setting the threshold. The linear regression model operator uses the multivariate linear regression algorithm as a basic tool for data training. The model training operator is a tool for executing model training.

7. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 5, characterized in that: The health evaluation of the auxiliary equipment health status monitoring AI model includes: The evaluation is carried out using the three evaluation index dimensions of MSE, MAE and R2, with the calculation results of MSE, MAE and R2 as the evaluation criteria. The lower the values ​​of the MSE and MAE calculation results indicators, the better the performance of the auxiliary equipment health status monitoring AI model; the higher the value of the R2 calculation result indicator, the better the performance of the auxiliary equipment health status monitoring AI model.

8. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to any one of claims 1, wherein: The AI+mechanism early warning model includes: Using the constructed AI model, we fit and mine the variable operating condition health operation relationship of the core monitoring parameters of the equipment, predict and output the upper and lower limit thresholds of the parameter operation under real-time conditions, and then perform logical operations and comparisons with the real-time parameter values. If the real-time value exceeds the predicted threshold, it will issue a warning and complete the health monitoring of the auxiliary equipment status; Logic operation components include AND, OR, NOT, R / S flip-flop, M / N, and signal hold.

9. The method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to claim 8, characterized in that: The auxiliary equipment status includes: Divide the equipment operating status according to the operating range of the auxiliary equipment unit consumption; When the coal mill unit consumption is greater than or equal to 4.8kWh / t and less than or equal to 8.2kWh / t, the equipment is in normal operation; When the coal mill unit consumption is less than 4.8kWh / t, it means that the coal mill unit consumption is low, and the coal mill unit consumption low warning operation is executed; when the coal mill unit consumption is greater than 8.2kWh / t, it means that the coal mill unit consumption is high, and the coal mill unit consumption high warning operation is executed.

10. A health status monitoring system for auxiliary equipment in a thermal power plant based on polynomial regression, characterized by: It includes a data acquisition and typical characteristic parameter module (100), a data pre-processing and model building module (200), and an equipment status monitoring and early warning module (300); The data acquisition and typical characteristic parameter module (100) includes a data acquisition module (101) and a typical characteristic parameter module (102). The data acquisition module (101) realizes real-time data acquisition through the Internet of Things platform. The data acquisition supports multiple acquisition modes such as DCS data interface, SIS system data interface and real-time database. The acquisition of historical data and real-time data of auxiliary equipment is realized by means of the Internet platform and TrendDB real-time database. The typical characteristic parameter module (102) combines the experience of operation experts to determine the core parameters representing the variable working condition characteristics of the auxiliary equipment. It further fully combines the operation and economic characteristics of the equipment. From the perspective of economy, some auxiliary equipment selects the unit consumption of the equipment as the key state parameter. From the perspective of safe operation, the correlation analysis algorithm is used to eliminate the redundant equipment measurement points with coherence, and the typical operation characteristic parameters with high correlation are mined to realize accurate and efficient state monitoring of the auxiliary equipment. The data preprocessing and model building module (200) includes a data preprocessing module (201) and a model building module (202). The data preprocessing module (201) is based on an Internet platform and uses an edge layer to perform equipment historical data quality judgment through a variety of methods such as histogram analysis, data fluctuation judgment, and unreasonable value range judgment, and then uses component operators to achieve data cleaning and improve the quality of training data. The model building module (202) includes a multiple linear regression AI model. The multiple linear regression AI model uses a visual modeling tool to quickly build a polynomial regression AI model of 6 major auxiliary equipment in a "drag and drop" manner based on the historical operation data of the past year. According to the floating trend of the historical data of the equipment characteristic parameters, the upper and lower floating ranges of the parameters are set, and finally a healthy range data set of the core parameters of the equipment is output; The equipment status monitoring and early warning module (300) comprises an equipment status monitoring module (301) and an equipment early warning module (302). The equipment status monitoring module (301) compares the health interval data set output by the constructed multivariate linear regression AI model with the real-time data collected by the equipment. If the data set is within the range of the health interval data set, it is a normal state; if the data set is beyond the range of the health interval data set, it is an unhealthy state. The equipment early warning module (302) uses the conclusion output by the equipment status monitoring module (301) to issue a high warning for the equipment in the unhealthy state if the data exceeds the range, and a low warning for the data exceeds the range.

11. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the processor implements the steps of a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a method for monitoring the health status of auxiliary equipment in a thermal power plant based on polynomial regression according to any one of claims 1 to 9 are implemented.

Images