An enhanced fault diagnosis method for fuel cell system using a kernel extreme learning machine optimized with improved sparrow search algorithm

Proton exchange membrane fuel cells (PEMFC) have a broad development prospect in the fields of vehicles, drones and ships due to their high efficiency and cleanliness. However, the problems of insufficient reliability and durability have severely restricted their industrialization process. To improve the safety, reliability and durability of fuel cell system, a fault diagnosis method that combined kernel principal component analysis (KPCA) with an improved sparrow search algorithm (ISSA) and an optimized kernel extreme learning machine (KELM) was proposed in this study. Firstly, KPCA is utilized to extract nonlinear features from fault indicators and obtain the fault feature vector of the fuel cell system. Then, by incorporating logistic mapping and Cauchy Gaussian mutation strategies to improve the Sparrow Search Algorithm (SSA), ISSA was used to optimize the kernel parameters and regularization coefficient in KELM. The experimental results show that the KPCA-ISSA-KELM method for normal conditions, hydrogen leakage and membrane drying are 100%, 98.5% and 100%, respectively, with an overall accuracy of 99.5% and an operation time of 0.97s. The diagnostic accuracy of the proposed method is 10.4%, 5.7%, 4.8%, 4.2%, 3.0%, 1.8% higher than support vector machine (SVM), back propagation neural network (BPNN), KELM, genetic algorithm-based KELM (GA-KELM), particle swarm optimization-based KELM (PSO-KELM) and SSA-KELM, respectively, and the operation time is only slightly higher than that of the SVM model and KELM model.