Real-time Fault Detection And Diagnosis Of Automatic Shift Control System

With the extensive application of vehicle automatic transmission, fault diagnosis as akey technology of automatic transmission, can effectively improve the maintenanceefficiency and vehicle driving safety. Specifically, the real-time fault detection anddiagnosis technology of Automatic Shift Control System (ASCS) are studied from therespects of theoretical analysis and experimental research with a heavy duty off-roadvehicle. The main contents are summarized as follows:AMT architecture is divided from the perspective of the hardware and software, andthe mapping relationship between the different levels of the software/hardware system isestablished. Based on the AMT system analysis, the terms of fault definition, faultdiagnostic purposes, fault propagation characteristics and fault diagnosability are studied.Through the behavioral characteristics of the ASCS, the fault prior knowledge and thediagnostic methods of the system different behavior types are obtained, and a real-timefault detection and diagnosis strategy of the ASCS is proposed.Based on the information prioritization of the system behavior types, the ASCSsteady-state behaviors are divided into different types to reduce the difficulty of faultdiagnosis and improve the fault diagnosability. Under the driver intent of the steady-statebehaviors, signal time-domain/frequency-domain analysis, signal collection mode andsignal transmission protocol are adopted to achieve fault diagnosis function. For thestationary random signals of the ASCS steady-state behaviors, the PCA theory is used toanalyse faultless historical data for fault detection and diagnosis of the stationary randomsignals. As for the non-stationary random signals under the steady-state behaviors, theredundancy relationship between the system signals is used to propose the fault detectionand diagnosis strategy based on the credible principle, the principle compatible withmultiple faults and the probability principle.According to the physical system response behaviors of the automatic shift controlsystem, first, the physical system response behaviors are divided into six sub-behaviortypes. And the logical combination relationship between the different sub-behaviors is analyzed. Then, the mathematical models of the transmission system and the shift controlsystem are established. The SDG theory is used for the mathematical model integration toobtain the SDG fault diagnosis model of the physical system response behaviors, and theexpert knowledge base.Finally, through the summary of the ASCS single behavior (steady-state behavior andphysical system response behavior), the fault diagnosis expert knowledge base is listed. Bythe depth study of the coupling relationship between the steady-state behaviors and thephysical system response behaviors, the fault combined diagnosis strategy is proposed andthe expert knowledge is obtained, which can effectively improve the accuracy of faultdiagnosis of the automatic shift control system.