Artificial Intelligent Techniques for Electric and Hybrid Electric Vehicles

12A New Approach for Flux Computation Using Intelligent Technique for Direct Flux Oriented Control of Asynchronous Motor

A. Venkadesan¹*, K. Sedhuraman², S. Himavathi³ and A. Chitra⁴

¹ NIT Puducherry, Karaikal, India

² MVIT Puducherry, Puducherry, India

³ Pondicherry Engineering College, Puducherry, India

⁴ EPE Department, SELECT, VIT, Vellore, India

Abstract

The accurate estimation of magnitude and angle of flux is vital for good performance of Direct Flux/Field Oriented Control (DFOC) of Asynchronous Motor (AM)/Induction Motor (IM). For this control, accurate value of flux plays a major role. The current equations (CE) of the motor estimate flux with stator current and rotor speed as the inputs without the need of difficult stator voltage PWM voltage measurement. Hence CE can be comfortably used for flux estimation. But CE majorly depends on the rotor resistance (R_r) and varies during motor operation. This leads to significant error in the flux estimation. To address this problem, an intelligent approach namely neural network is employed. A novel Neural Network (NN) approach for flux estimation is proposed in this paper. The proposed approach uses stator current and rotor speed as the inputs similar to CE. The proposed NN based estimator is shown to handle rotor resistance variation problem as compared to current model-based flux estimator through MATLAB simulation.

Keywords: Flux estimator, current model, intelligent technique, neural network, induction motor, direct field ...