Research on engine fault diagnosis technology based on ANFIS

Aim to the problem of engine fault diagnostic accuracy, research an engine fault diagnosis technologies based on adaptive neuro-fuzzy inference system. The method is proposed that using known input and output data, construct a fuzzy inference system, internal systems adjust the membership function using a least squares method based on hybrid learning algorithm and reverse the spread of the combination by four major state test engine (including a normal state, the stator short-circuit fault, the rotor rupture fault and the rotor eccentricity fault) of experimental data, The results shows that the outputs of test is basic agreement with the expected outputs. The algorithm is proposed and it is used for fault diagnosis of asynchronous engine with high diagnostic accuracy, it can overcome the traditional engine fault diagnosis method has big error, for a long time and other shortcomings.