Research on fault diagnosis of non-stationary random signals under ASCS's steady state condition

A multiple fault detection and diagnosis strategy of non-stationary random signals is presented for automatic shift control system (ASCS), which is based on the redundancy relationship between signals under the steady-state condition. The design principles of fault diagnosis strategy are put forward based on the characteristics of ASCS non-stationary random signals and the redundancy relationship between them. These principles, including credibility principle, fault compatible principle and probability principle, are the design guidelines of fault diagnosis strategy. According to the accuracy and efficiency of fault diagnosis results, the results are divided into three types. Then the fault detection and diagnosis strategy of non-stationary random signals is put forward on the basis of prior knowledge under the steady state condition of ASCS, which can realize the single fault or multiple fault diagnosis of engine speed, transmission input speed, vehicle speed and gear signal. Man-made fault is injected through the real vehicle test process to detect the integrity and validity of fault diagnosis strategy. The car data and test results show that the method is effective and practical.