摘要
针对固体氧化物燃料电池(SOFC)系统的高维、非线性及多工况等特点,提出了基于PCA-HSVM的SOFC故障识别策略。首先,利用PCA提取故障特征信息,然后在降维后的特征空间里采用层级法构建HSVM多分类模型。实验结果表明:PCA-HSVM能够更加准确、快速地识别故障类型。
Considering SOFC( solid oxide fuel cell) system's characteristics of high dimension,nonlinearity,and multiple working conditions,the PCA-HSVM-based SOFC fault diagnosis strategy was proposed. Firstly,having PCA adopted to extract fault feature information and then having the hierarchical approach used to establish HSVM multi-classification model in the reduced dimension space. Experimental results show that,the PCA-HSVM can identify fault types more accurately and rapidly.
作者
秦超
李双宏
杨煜普
QIN Chao, LI Shuang-hong, YANG Yu-pu(School of Electronic Information and Electrical Engineering, Shanghai Jiaotong Universit)
出处
《化工自动化及仪表》
CAS
2018年第8期611-616,共6页
Control and Instruments in Chemical Industry
基金
国家自然科学基金项目(51777122)