Application Research of Fault Diagnosis in Conventional Island of Nuclear Power Plant Based on Support Vector Machine

被引：1

作者：

Li, Heng ^{[1
]}

Lan, Nian-Wu ^{[1
]}

Huang, Xin-nian ^{[1
]}

机构：

[1] China Nucl Power Engn Co Ltd, State Key Lab Nucl Power Safety Monitoring Techno, Shenzhen 518172, Guangdong, Peoples R China

来源：

NUCLEAR POWER PLANTS: INNOVATIVE TECHNOLOGIES FOR INSTRUMENTATION AND CONTROL SYSTEMS (ISNPP 2019) | 2020年 / 595卷

关键词：

Support vector machine; Machine learning; Nuclear power plant conventional island; Fault diagnosis;

D O I：

10.1007/978-981-15-1876-8_31

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

From the historical data of pump device in conventional island of a certain nuclear power plant, the dataset used for machine learning is selected and established by whether the device is fault. Then the dataset is divided into the training set and the test set. Relying on the powerful machine learning library of Python language, the support vector machine model is constructed by programming. After selecting the appropriate kernel function and hyperparameters, the fault diagnosis accuracy of the support vector machine model on the test set reaches a high level. The generalization ability of the model is strong, which proves the model can be used as an auxiliary means for the fault diagnosis in conventional island of nuclear power plant.

引用

页码：304 / 312

页数：9