FUEL CELL FAULT DIAGNOSIS METHOD AND SYSTEM

The present disclosure provides a fuel cell fault diagnosis method and system. The method includes: determining key parameters for temporal feature and spatial feature extractions according to the amount of information contained in all sensor parameters in a fuel cell to be diagnosed; obtaining a target intrinsic mode function by performing complementary ensemble empirical mode decomposition on a temporal feature matrix; obtaining a high-dimensional mapping vector by processing spatial feature data using a kernel principal component analysis; and obtaining a spatial feature to be synchronized by processing the high-dimensional mapping vector according to a preset confidence interval, obtaining a temporal feature to be synchronized by processing a target decomposition result according to a lumped fuzzy entropy operation, fusing the spatial feature to be synchronized and the temporal feature to be synchronized, and obtaining a fault diagnosis result by inputting a synchronous spatial-temporal feature into a preset diagnosis model.