Fault Diagnosis of PEMFC Stack Based on PSO-DBN

As an important means of fault diagnosis for proton exchange membrane fuel cell (PEMFC), data-driven method can make accurate fault diagnosis by training a large number of fault sample data. Aiming at the problem that the more the dimension of the sample, the longer the learning time, this paper proposes a dimension reduction algorithm based on principal component analysis (PCA), which maps the multi-dimensional original data to the low-dimensional new data, and greatly accelerates the training efficiency under the premise of ensuring the reflection of the fault. At the same time, the fault diagnosis method based on the traditional machine learning model cannot accurately classify the data set of multi-dimensional features generated by the complex system of PEMFC stack, which leads to the low accuracy of fault diagnosis. The particle swarm optimization deep belief network (PSO-DBN) algorithm is designed to realize the PEMFC fault diagnosis method with high diagnostic accuracy. The experimental results show that the accuracy of the fault diagnosis algorithm based on deep learning proposed in this paper can reach 99.7% for the test set, and the efficiency and accuracy of fault diagnosis are better than traditional machine learning fault diagnosis algorithms such as Back propagation (BP) and support vector machine (SVM).