A General Approach for Fault Identification in SOFC-based Power Generation Systems

作者：Zhang, Zehan^*; Li, Shuanghong; Yang, Yupu

来源：2018 Annual American Control Conference, ACC 2018, 2018-06-27 To 2018-06-29.

DOI：10.23919/ACC.2018.8430954

The design of fault identification method in fuel cells(FCs)-based systems remains a key challenge. This paper proposes a general data-based fault identification method using deep learning(DL). Because of the great success of deep learning in a variety of practical applications, it has recently become the most popular approach in machine learning. Deep learning employs a hierarchical structure to learn the inherent features in systems layer by layer. The proposed DL-based fault identification method can be structured into three stages: first, transforming the fault identification problem into classification problem, defining the normal and fault state, and preprocessing the original data;second, developing the DL model for fault identification using the preprocessed data and the predefined state;third, validating the DL model with testing dataset. The proposed method is applied to two different size and multiple fault types datasets generated by the SOFC system. And the testing results show that the proposed method can accurately predict the fault type outperforming the other two classification methods. In addition, since the proposed method requires only data, it is easy to migrate to other types of FC-based systems.

• 出版日期2018
• 单位上海交通大学