Research On Inertia Identification Algorithm Of Permanent Magnet Synchronous Motor

Permanent magnet synchronous motor is widely used in high performance controlfield with its the advantages of high reliability, high efficiency, high power density, goodlow-speed performance, small rotor moment of inertia and simple control system. Becausethe mechanical properties of the permanent magnet synchronous motor servo system isaffected by the moment of inertia, if the moment of inertia can be real-time identified andthe control system can be adjusted according to the result of recognition, control effect ofservo system can be improved.0n the basis of studying the common inertia identification algorithm, in this thesis theleast squares inertia identification algorithm, the model reference adaptive inertiaidentification algorithm and gradient correction inertia identification algorithm are studied.The inertia identification simulation models of least squares algorithm, model referenceadaptive algorithm and gradient correction algorithm are established on Matlab platform.This thesis deeply analyzed the effect of the selection of forgetting factor on the leastsquare algorithm; the effect of the selection of adaptive gain factor on the model referenceadaptive algorithm; the identification results of model reference adaptive algorithm andinfluence of a/c on the the gradient correction algorithm; with the comparison ofsimulation results, this thesis gave the suggestion of the use of each algorithm.According to the defects of the above three identification algorithm, this thesis hasproposed this simple correction algorithm. And simulation tests on the three correctionalgorithm were done on the platform of matlab. The results show that every correctionalgorithm can effectively improve the identification results of the original algorithm andreduce the difficulty of the selection of the original algorithm.