Study on efficiency optimization control of induction motor drive system

This paper presents a new approach for efficiency optimization of a vector controlled induction motor drive. From the expression of the induction motor losses, we can see that the nonlinear relationship of the motor losses with the electromagnetic torque and rotor speed exists, and the total losses are affected by parameter variations of the induction motor. Therefore, it is difficult to attain satisfactory performance by using traditional methods. Neural network has self-organization and self-study ability, which is suitable to solve nonlinear problems with parameter variations. In order to minimize losses to obtain optimal rotor flux, a new algorithm incorporating neural network with layer-to-layer prediction and impending is proposed This method produces suitable rotor flux to fulfill optimal efficiency in complex operation condition. Simulation results show an obvious improvement in efficiency optimization.