摘要详情

ID / 提交时间

18 / 2024-05-30 15:54:02

标题

ABNORMAL MONITORING OF SENSOR MEASUREMENT INFORMATION UNDER MULTIPLE WORKING CONDITIONS IN NUCLEAR POWER PLANTS

关键字

Abnormal Monitoring; Dynamic Correlation; Fuzzy C-means Clustering; Gaussian Mixture Model

主题及专题

TRACK 2: OPERATION & PERFORMANCE

状态

终稿

作者

Hongyuan Chen / Harbin Engineering University

Minjun Peng / Harbin Engineering University

Hang Wang / Harbin Engineering University

Chen Li / Harbin Engineering University

摘要

With the optimization of energy structure and the improvement of environmental protection requirements, nuclear energy, as an important clean energy, occupies an important position in the global energy structure. However, the complex operation process and potential high risk of nuclear power plants put forward strict requirements for real-time monitoring. As one of the key technologies for the safe operation of nuclear power plants, multi-condition condition monitoring is not only related to the stable operation of the power grid, but also an important way to improve the economic benefits of nuclear power plants, reduce operating costs and ensure public safety. Therefore, the research on multi-condition condition monitoring of nuclear power plants has important academic value and practical significance. In view of the difficulty of condition monitoring caused by high-dimensional, strongly coupled and fast time-varying data in nuclear power plants, an anomaly monitoring model for multiple working conditions is proposed. Firstly, in order to solve the problem that the accuracy of the existing condition monitoring model is not high under variable working conditions, it is divided into multiple distributed monitoring units based on the dynamic correlation between measurement parameters, and a plurality of monitoring task pieces is established based on historical operation data, and then the static and dynamic characteristics of each unit are obtained by combining the data reorganization strategy. Secondly, based on the obtained features, a multi-layer progressive monitoring model based on fuzzy C-means clustering and Gaussian mixture model fusion was proposed to realize anomaly monitoring under multiple working conditions and improve the accuracy of anomaly identification. Finally, based on the experimental bench and the integrated simulator, the normal and abnormal data under multiple working conditions of the nuclear power plant are obtained to complete the model verification, and the results show that the above model can accurately and effectively complete the abnormal monitoring of sensor measurement signals. The proposed model can alleviate the pressure of operators to a certain extent and reduce the accidents caused by human error. At the same time, through the in-depth study of data characteristics, it provides a feasible method for condition monitoring under variable working conditions.