关键词：汽车；故障检测；判别分析
摘 要：In this paper, fisher discriminant analysis (FDA) is used for fault isolation and diagnosis in rail vehicle suspension systems. The suspension systems are equipped with acceleration sensors in the corners of the car body and the two bogies. The faults considered are the lateral damper faults and the lateral spring faults in suspension systems. FDA provides an optimal projection space on the basis of the training data including the fault data and normal data to classify the test data. A vehicle model is built by SIMPACK/MATLAB software with real parameters to obtain the simulation data and the effectiveness of the proposed method is demonstrated by simulation.