CN-122022767-A - Nuclear power station operation fault prediction system and prediction method

CN122022767ACN 122022767 ACN122022767 ACN 122022767ACN-122022767-A

Abstract

The invention provides a prediction system and a prediction method for operation faults of a nuclear power station, wherein the prediction system comprises a data acquisition module, a data prediction module, an abnormality analysis module and a fault diagnosis module, wherein the data acquisition module is used for acquiring monitoring data of each monitoring device of the nuclear power station and acquiring corresponding measured data calculated according to the monitoring data, the data prediction module is used for inputting the monitoring data into a digital twin model of the nuclear power station to obtain corresponding prediction data, the abnormality analysis module is used for comparing the measured data with the corresponding prediction data to obtain a corresponding abnormal characteristic mode, and the fault diagnosis module is used for matching the abnormal characteristic mode with a preset fault mode library to obtain a corresponding abnormal fault type. The prediction system and the prediction method for the operation faults of the nuclear power station provided by the invention are used for solving the technical problem that the active early warning and the accurate decision of the nuclear power station are difficult to realize.

Inventors

PAN JINGBIN
YUAN CHAO
HUI ZHIGUANG
HUANG YONGCHENG
LIU DONGBO
SHI SHUJIAN

Assignees

中广核清洁能源科技(上海)有限公司

Dates

Publication Date: 20260512
Application Date: 20260119

Claims (10)

1. A system for predicting an operational failure of a nuclear power plant, comprising: The data acquisition module is used for acquiring monitoring data of each monitoring device of the nuclear power station and acquiring corresponding measured data obtained by calculation according to the monitoring data; The data prediction module is used for inputting the monitoring data into a digital twin model of the nuclear power station to obtain corresponding prediction data; the abnormality analysis module is used for comparing the measured data with corresponding predicted data to obtain a corresponding abnormality characteristic mode; and the fault diagnosis module is used for matching the abnormal characteristic mode with a preset fault mode library to obtain a corresponding abnormal fault type.
2. The nuclear power plant operational fault prediction system of claim 1, wherein the anomaly analysis module is configured to: comparing the measured data with corresponding predicted data to obtain corresponding characteristic data; And identifying abnormal characteristic data, and obtaining a corresponding abnormal characteristic mode according to the abnormal characteristic data.
3. The nuclear power plant operational fault prediction system of claim 2, wherein the anomaly analysis module is configured to: Comparing the measured data with corresponding predicted data to obtain a corresponding residual sequence; And processing the residual sequence to obtain corresponding characteristic data.
4. A prediction system of nuclear power plant operational faults as claimed in claim 3, in which the anomaly analysis module is for: segmenting the residual sequence based on a preset sliding time window; and calculating the mean value, variance, skewness and kurtosis of the residual data in each time window as corresponding characteristic data.
5. The nuclear power plant operational fault prediction system of claim 2, wherein the anomaly analysis module is configured to: Comparing the characteristic data with a corresponding threshold value, wherein when any one of the variation amplitude, the duration direction and the statistical distribution of the characteristic data exceeds the corresponding threshold value, the characteristic data is judged to have abnormality; And determining a corresponding abnormal characteristic mode according to the change mode of the characteristic data with the abnormality.
6. A prediction system for an operational fault of a nuclear power plant as defined in claim 1, further comprising a fault prediction module for calculating a prediction probability of a particular fault occurring in the monitoring device within a specified period of time in the future based on the monitored data of the monitoring device.
7. The nuclear power plant operational fault prediction system of claim 6, wherein the fault prediction module is configured to: Extracting corresponding characteristic parameters according to the monitoring data of the monitoring equipment; And inputting the working parameters, the monitoring data and the corresponding characteristic parameters of the monitoring equipment into a predictive sub-model in the digital twin model to obtain the predictive probability of the specific fault of the monitoring equipment in a future designated period, wherein the predictive sub-model calculates the predictive probability based on a Bayesian updating algorithm or a machine learning classification algorithm.
8. The system for predicting an operational failure of a nuclear power plant of claim 1, further comprising a failure inference module configured to obtain an inference event report and a corresponding equipment maintenance scheme for a failure of the monitoring equipment affecting the nuclear power plant when the monitoring equipment is determined to be failed.
9. The nuclear power plant operational fault prediction system of claim 8, wherein the fault deduction module is configured to: Acquiring a monitoring device determined to be invalid; Inputting monitoring parameters of the monitoring equipment which is judged to be invalid and the topological position of the monitoring equipment in the nuclear power station into the digital twin model to obtain a deduction event report of the influence of the failure of the monitoring equipment on the nuclear power station; And matching the deduction event report with a preset failure scene knowledge base, and determining an equipment maintenance scheme based on a matching result.
10. A method for predicting an operational failure of a nuclear power plant, comprising: Acquiring monitoring data of each monitoring device of the nuclear power station, and acquiring corresponding measured data obtained by calculation according to the monitoring data; inputting the monitoring data into a digital twin model of the nuclear power station to obtain corresponding prediction data; Comparing the measured data with corresponding predicted data to obtain a corresponding abnormal characteristic mode; and matching the abnormal characteristic mode with a preset fault mode library to obtain a corresponding abnormal fault type.

Description

Nuclear power station operation fault prediction system and prediction method Technical Field The invention relates to the field of nuclear power, in particular to a prediction system and a prediction method for operation faults of a nuclear power station. Background At present, the maintenance and management of domestic nuclear power stations mainly depend on periodic maintenance and manual inspection. Periodic inspections inspect and maintain equipment through a fixed period, while preventing some failures, lack of real-time may result in a delay in failure discovery. Manual inspection is limited by personnel experience and work intensity, and is difficult to cover the potential risks of complex equipment systems. In addition, excessive reliance on regular maintenance may result in frequent disassembly and assembly of the equipment, rather shortening its service life and increasing operational and maintenance costs. At present, the nuclear power station lacks the capability of real-time monitoring of equipment states and dynamic deduction of fault scenes, active early warning and accurate decision making are difficult to realize, so that fault discovery is delayed, operation and maintenance efficiency is low, and the requirements of high safety and stable operation of the nuclear power station are difficult to meet. Therefore, there is a need for improvement. Disclosure of Invention The invention provides a prediction system and a prediction method for operation faults of a nuclear power station, which aim to solve the technical problem that active early warning and accurate decision making of the nuclear power station are difficult to realize. The invention provides a prediction system for operation faults of a nuclear power station, which comprises the following components: The data acquisition module is used for acquiring monitoring data of each monitoring device of the nuclear power station and acquiring corresponding measured data obtained by calculation according to the monitoring data; The data prediction module is used for inputting the monitoring data into a digital twin model of the nuclear power station to obtain corresponding prediction data; the abnormality analysis module is used for comparing the measured data with corresponding predicted data to obtain a corresponding abnormality characteristic mode; and the fault diagnosis module is used for matching the abnormal characteristic mode with a preset fault mode library to obtain a corresponding abnormal fault type. In an embodiment of the present invention, the anomaly analysis module is configured to: comparing the measured data with corresponding predicted data to obtain corresponding characteristic data; And identifying abnormal characteristic data, and obtaining a corresponding abnormal characteristic mode according to the abnormal characteristic data. In an embodiment of the present invention, the anomaly analysis module is configured to: Comparing the measured data with corresponding predicted data to obtain a corresponding residual sequence; And processing the residual sequence to obtain corresponding characteristic data. In an embodiment of the present invention, the anomaly analysis module is configured to: segmenting the residual sequence based on a preset sliding time window; and calculating the mean value, variance, skewness and kurtosis of the residual data in each time window as corresponding characteristic data. In an embodiment of the present invention, the anomaly analysis module is configured to: Comparing the characteristic data with a corresponding threshold value, wherein when any one of the variation amplitude, the duration direction and the statistical distribution of the characteristic data exceeds the corresponding threshold value, the characteristic data is judged to have abnormality; And determining a corresponding abnormal characteristic mode according to the change mode of the characteristic data with the abnormality. In an embodiment of the present invention, the prediction system further includes a fault prediction module, where the fault prediction module is configured to calculate, according to the monitoring data of the monitoring device, a prediction probability that the monitoring device generates a specific fault in a specified future period. In an embodiment of the present invention, the fault prediction module is configured to: Extracting corresponding characteristic parameters according to the monitoring data of the monitoring equipment; And inputting the working parameters, the monitoring data and the corresponding characteristic parameters of the monitoring equipment into a predictive sub-model in the digital twin model to obtain the predictive probability of the specific fault of the monitoring equipment in a future designated period, wherein the predictive sub-model calculates the predictive probability based on a Bayesian updating algorithm or a machine learning classification algorithm. In an embo