Nuclear power plant equipment monitoring data clustering processing method and electronic equipment

Abstract

The invention relates to a nuclear power plant equipment monitoring data clustering processing method and electronic equipment. The method comprises the following steps: S1, acquiring monitoring data; S2, standardizing the monitoring data to generate a to-be-clustered data set; S3, clustering all to-be-clustered data in the to-be-clustered data set to obtain initial class data; S4, performing class integration by using the initial class data to obtain a clustering result; S5, judging whether the clustering result meets the clustering target requirement or not; S6, if so, stopping clustering; and S7, if not, continuing to carry out clustering. The invention can effectively adapt to clustering analysis of state monitoring historical data in the steady-state operation process of nuclear power plant equipment, and can effectively adapt to clustering analysis of high-dimensional data, and the obtained clustering result is continuous and compact in space and high in calculation efficiency.

Description

technical field [0001] The invention relates to the technical field of nuclear power plant equipment status monitoring data processing, and more specifically, relates to a nuclear power plant equipment monitoring data clustering processing method and electronic equipment. Background technique [0002] At present, there are tens of thousands of online monitoring data measuring points for a 900MW nuclear power unit in China. The historical data of equipment monitoring can exceed 30 years at most, and the amount of data is extremely large. Operation status monitoring service is an extremely important key technology in the application of nuclear power equipment management big data. Relying on traditional statistical analysis or expert identification methods cannot adapt to the processing of huge data, and cannot discover the laws hidden in massive equipment monitoring data. Therefore, an unsupervised learning algorithm is needed to realize automatic clustering of huge historical...

AI Technical Summary

Problems solved by technology

Relying on traditional statistical analysis or expert identification methods cannot adapt to the processing of huge data, and cannot discover the laws hidden in massive equipment monitoring data. Therefore, an unsupervised learning algorithm is needed to realize automatic clustering of huge historical data. Identify patterns hidden in historical data

[0003] Because equipment monitoring needs to avoid missing alarms and maintain an extremely low false alarm rate, the current classic clustering algorithm equipment management and monitoring applications are obviously not suitable

Images