In the existing SOFC system, because of the high temperature isolation of the hot box, it is difficult to obtain the situation of the SOFC stack accurately. Therefore, the key problem of real-time judgment of the SOFC system operation state is to develop a state diagnosis method. Therefore, this paper intends to use strong classifier to evaluate the state of SOFC system, which is consist of multi weak classifiers and called AdaBoost algorithms. The algorithms can achieves efficient classification, which provides a powerful guidance for the overall performance evaluation of the system.
Keywords Solid oxide fuel cells (SOFC), Adaptive boosting (AdaBoost), Fault diagnosis, System analysis