On-Line Diagnosis of the Motor Circuit Based on RBF-CMGA

Based on the syncretic theory of the radial basal function neural network (RBFNN) and the contractive mapping genetic arithmetic (CMGA), an on-line diagnostic strategy on the motor circuit fault for vehicles was proposed. The fault on the motor drive circuit were simulated by PSpice, the sampling data were regarded as the inputs of RBFNN, and the faults codes on the components of the drive circuit were regarded as the output of RBFNN, the diagnostic strategy on the motor drive circuit was educated and studied by RBF-CMGA. The on-line diagnostic tests on the fault, such as, short circuit and open circuit of on the drive circuit components were made in the cold-starting engine laboratory by the RBF-CMGA syncretic theory. The test results that the diagnostic accuracy could also reach 94.3%, the error rate is 0.58%, and the rejective rate is 0.15%, which was testified that the diagnostic strategy could carry out the accurate diagnosis and the orientation of the motor drive circuit malfunction, which completely satisfied the motor started successfully and reliably.