The Neural Network Inverse Control Method of Induction Motor Based on Multiscalar Model

The decoupling and linearisation (D&L) of induction motor (IM) is an important approach to improve the control performance further. The multiscalar model of IM owns many advantages. So, based on the multiscalar model of IM, the analytic inverse system (ANIS) theory was used to analyze the invertibility of the IM system, and the analytic inverse control (ANIC) law was deduced. For the IM with parameters varying and external disturbance, the obtained D&L by ANIC is destroyed. So the neural network inverse system (NNIS) theory was adapted to design the NNIS of the IM, that is, the ANIS was replaced with the NNIS in order to weaken the couple between rotor speed and rotor flux, thus the high static and dynamic control performance of IM can be obtained. At last, the simulation was done to test that the proposed structure is valid and it is more robust than that of ANIC.