Fault Diagnosis of PEMFC Systems Based on Decision-making Tree Classifier

For purpose of solving the problem of fault diagnosis of the evaporatively cooled fuel cell system, a novel fast fault diagnosis method for an evaporatively cooled fuel cell system based on decision-making tree classifier is developed in the paper. The normalization method is utilized to filter the original datum. The decision-making tree classifier is used to classify the preprocessed data. It can effectively improve the diagnostic accuracy of the model. Example analysis shows that the novel approach can rapidly recognize two health states of membrane drying fault and hydrogen leakage fault. The diagnostic accuracy of the algorithm is 98.50%. The method proposed in this paper is suitable for online failure identification of evaporatively cooled fuel cell systems with large data samples and multi data dimensions.