阅读说明：本技术 一种核电站主泵状态监测诊断方法及系统 (Nuclear power station main pump state monitoring and diagnosing method and system ) 是由 赵金栋 任成宾 郭英端 曹福森 于庆海 自明 王立峰 潘爱兵 刘汝玉 彭楠 赵德 于 2020-12-28 设计创作，主要内容包括：本发明涉及一种核电站主泵状态监测诊断方法和系统,所述的方法包括以下步骤：S1：数据收集并确定监测点的步骤,包括：读取主泵的文件资料数据,包括反应堆冷却技术规范数据、反应堆安全分析数据、主泵运行数据、主泵图纸数据；读取主泵仪表数据和监测点编码；S2：针对主泵建立多元非线性回归模型的步骤。(The invention relates to a method and a system for monitoring and diagnosing the state of a main pump of a nuclear power station, wherein the method comprises the following steps: s1: the method comprises the steps of collecting data and determining monitoring points, and comprises the following steps: reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data; reading main pump instrument data and monitoring point codes; s2: and establishing a multiple nonlinear regression model for the main pump.)

1. A method for monitoring and diagnosing the state of a main pump of a nuclear power station is characterized by comprising the following steps:

s1: the method comprises the steps of collecting data and determining monitoring points, and comprises the following steps:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes;

s2: the step of establishing a multiple non-linear regression model for the main pump comprises the following steps:

s21: selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

s22: adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group;

s23: inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

s24: dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

2. The method for monitoring and diagnosing the condition of the main pump of the nuclear power plant as claimed in claim 1, wherein in the step S22, irrelevant monitoring points of the monitoring point group are screened and removed by using a correlation analysis means.

3. The method for monitoring and diagnosing the condition of the main pump of the nuclear power plant as claimed in claim 1 or 2, further comprising:

s3: and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

4. The method for monitoring and diagnosing the condition of the main pump of the nuclear power plant as recited in claim 3, further comprising:

s4: the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

5. The method for monitoring and diagnosing the condition of the main pump of the nuclear power plant as recited in claim 4, further comprising:

s5: and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

6. A nuclear power station main pump condition monitoring and diagnosing system is characterized by comprising:

a data collection and monitoring point determination module, wherein:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes;

establishing a multivariate nonlinear regression model module, wherein:

selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group;

inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

7. The system for monitoring and diagnosing the condition of a main pump of a nuclear power plant as claimed in claim 6, further comprising:

a data display module, wherein:

and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

8. The system for monitoring and diagnosing the condition of a main pump of a nuclear power plant as claimed in claim 7, further comprising:

the equipment index of the same type compares the module, in this module:

the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

9. The system for monitoring and diagnosing the condition of a main pump of a nuclear power plant as recited in claim 8, further comprising:

constructing a fault characteristic library module, wherein:

and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

Technical Field

The invention belongs to the technical field of state monitoring and diagnosis of rotating equipment, and particularly relates to a method and a system for monitoring and diagnosing the state of a main pump of a nuclear power station.

Background

A main heat transfer pump (a main pump for short) of a nuclear power station is important equipment of the nuclear power station, the current domestic mainstream nuclear power technology is a pressurized water reactor, and the main pump mainly comprises a shield pump and a shaft seal pump. The reactor is one of important pumps of a Reactor Coolant System (RCS), and has the main functions of driving medium circulation in a first loop to take away heat in the reactor and transferring the heat to a second loop through a steam generator so as to achieve core cooling and reactivity control.

The failure of the main pump under normal operation can cause shutdown, and the failure of the main pump under accident conditions can affect the safety of a reactor core. Therefore, the condition monitoring and diagnosis level of the main pump is related to safe and reliable operation of the whole nuclear power plant, and the economic efficiency of the nuclear power plant is also affected if the shutdown is caused.

The main pump monitoring research fields at home and abroad are most common in the field of equipment design and the field of operational reliability. In the field of equipment design, the design improvement of a main pump is focused, and the purposes of improving the running stability and the sealing reliability of the main pump and the like are achieved through the reasonable design of equipment hardware; in the field of operation reliability, a monitoring method, a fault mode and a mechanism analysis for mainly analyzing common faults of pumps to which the main pump belongs are adopted.

But the method is limited by less main pump measuring points and low key technology mastery degree, has less improvement and research on the state monitoring method of the produced main pump, and lacks a system which can be put into practical application to improve the monitoring level of the main pump. This is a disadvantage of the prior art.

In view of this, the invention provides a method and a system for monitoring and diagnosing the state of a nuclear power station main pump; it is very necessary to solve the above-mentioned defects existing in the prior art.

Disclosure of Invention

The invention aims to provide a method and a system for monitoring and diagnosing the state of a main pump of a nuclear power station, aiming at the defects in the prior art, so as to solve the technical problems.

In order to achieve the purpose, the invention provides the following technical scheme:

a method for monitoring and diagnosing the state of a main pump of a nuclear power station comprises the following steps:

s1: the method comprises the steps of collecting data and determining monitoring points, and comprises the following steps:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes, wherein the main pump instrument data is obtained by reading a distributed control system, and the monitoring point codes are output to a database of a power plant management area through the distributed control system to obtain;

s2: the step of establishing a multiple non-linear regression model for the main pump comprises the following steps:

s21: selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

s22: adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group;

s23: inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

s24: dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

Preferably, in step S22, irrelevant monitoring points in the monitoring point group are filtered and removed by a correlation analysis means; monitoring points irrelevant to the main pump are removed more accurately.

Preferably, the diagnostic method further comprises:

s3: the step of displaying, comprising:

and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

Preferably, the diagnostic method further comprises:

s4: the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

Preferably, the diagnostic method further comprises:

s5: and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

The invention also provides a system for monitoring and diagnosing the state of the main pump of the nuclear power station, which comprises the following components:

a data collection and monitoring point determination module, wherein:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes, wherein the main pump instrument data is obtained by reading a distributed control system, and the monitoring point codes are output to a database of a power plant management area through the distributed control system to obtain;

establishing a multivariate nonlinear regression model module, wherein:

selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group;

inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

Preferably, the diagnostic system further comprises:

a data display module, wherein:

and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

Preferably, the diagnostic system further comprises:

the equipment index of the same type compares the module, in this module:

the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

Preferably, the diagnostic system further comprises:

constructing a fault characteristic library module, wherein:

and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

The method has the advantages that the result is calculated in real time, the running state parameters are displayed in real time, and compared with the existing method for analyzing after-accident, the timeliness is improved, and engineers can conveniently master the performance condition of the equipment in time. And by adopting graphical display, compared with an accident analysis report, the method is more intuitive, and the data visualization effect is obvious. According to the main pump state monitoring and diagnosing method based on the regression model, early warning is carried out according to residual errors obtained by the regression model, possible faults are prompted, and the operation reliability and economy of the main pump are improved.

In addition, the invention has reliable design principle, simple structure and very wide application prospect.

Therefore, compared with the prior art, the invention has prominent substantive features and remarkable progress, and the beneficial effects of the implementation are also obvious.

Drawings

Fig. 1 is a flowchart of a method for monitoring and diagnosing the condition of a main pump of a nuclear power plant according to the present invention.

Fig. 2 is a schematic block diagram of a system for monitoring and diagnosing the condition of a main pump of a nuclear power plant according to the present invention.

The method comprises the following steps of 1-data collection and monitoring point determination module, 2-establishment of a multivariate nonlinear regression model module, 3-data display module, 4-same type equipment index comparison module and 5-establishment of a fault characteristic library module.

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings by way of specific examples, which are illustrative of the present invention and are not limited to the following embodiments.

Example 1:

as shown in fig. 1, the method for monitoring and diagnosing the condition of a main pump of a nuclear power plant according to the present embodiment includes the following steps:

s1: the method comprises the steps of collecting data and determining monitoring points, and comprises the following steps:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes, wherein the main pump instrument data is obtained by reading a distributed control system, and the monitoring point codes are output to a database of a power plant management area through the distributed control system to obtain;

s2: the step of establishing a multiple non-linear regression model for the main pump comprises the following steps:

s21: selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

s22: adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group; screening and removing irrelevant monitoring points of the monitoring point group by using a correlation analysis means; monitoring points irrelevant to the main pump are removed more accurately.

S23: inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

s24: dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

S3: and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

S4: the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

S5: and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

Example 2:

as shown in fig. 2, the system for monitoring and diagnosing a condition of a main pump of a nuclear power plant according to the present embodiment includes:

data collection and determination of monitoring points module 1, in which:

reading file data of a main pump, including reactor cooling technical specification data, reactor safety analysis data, main pump operation data and main pump drawing data;

reading main pump instrument data and monitoring point codes, wherein the main pump instrument data is obtained by reading a distributed control system, and the monitoring point codes are output to a database of a power plant management area through the distributed control system to obtain;

establishing a multivariate nonlinear regression model module 2, wherein:

selecting a time period at least comprising one refueling period as the establishment duration of the sample library;

adding monitoring points of the reactor cooling system as monitoring points into a monitoring point group, and then removing the monitoring points irrelevant to the main pump from the monitoring point group;

inquiring fault abnormal events of the main pump in the time period in the power plant work order and the state report order, and constructing the time period of the fault abnormal events into an abnormal working condition library;

identifying the time period of a main pump starting process and a main pump stopping process in the time period, and constructing a starting and stopping working condition library;

constructing a normal operation condition library in the residual time period;

constructing three sample libraries of an abnormal working condition library, a start-stop working condition library and a normal operation working condition library;

dividing a normal operation condition sample library according to data quantity, establishing a representative condition, carrying out modeling verification on a confirmed measuring point combination sample library to select a proper regression model, constructing a multiple nonlinear regression model, calculating by using a model algorithm of a support vector machine based on Euclidean distance to obtain an evaluation value of each measuring point under the normal operation condition library, and taking a result of subtracting the evaluation value from a data value farthest from the evaluation value as a threshold value of a residual error according to data of a 99.73% confidence interval of the representative condition in which each evaluation value is positioned.

A data display module 3 in which:

and (4) displaying the measured point value and the evaluated value of the main pump on a WEB display interface, and simultaneously displaying the residual error of the measured point, wherein if the residual error exceeds the calculated threshold, the measured point displays a red alarm state. And warning information is given to facilitate observation.

The same type equipment index compares the module 4, in this module:

the method comprises the steps of establishing a comparison analysis method of indexes of the same type of equipment, carrying out comparison analysis on the tendency of establishing the same index for main pumps under the same unit, and identifying whether the equipment is still in a healthy state or not when interference is generated by environmental factors or unknown factors.

Constructing a fault feature library module 5, wherein:

and constructing a fault feature library, adding main pump fault data of the same industry into the feature library, supporting fault addition and warehousing analyzed after parameter alarming, and prompting possible faults after subsequently generating similar alarming.

The above disclosure is only for the preferred embodiments of the present invention, but the present invention is not limited thereto, and any non-inventive changes that can be made by those skilled in the art and several modifications and amendments made without departing from the principle of the present invention shall fall within the protection scope of the present invention.