Nonlinear Fractional Order Model Identification of the Voltage Source Inverter Fed Induction Motor

Abstract

This paper presents a new nonlinear fractional-order model identification of a two-level voltage source inverter fed induction motor (VSI-IM). The nonlinearities present in inverter and IM cause the failure of the sensorless controls and model predictive controls, which mainly depends on the VSI-IM model. In the proposed work, an attempt has been made to identify the exact nonlinear behavior of the VSI-IM model using the PWM signal injection method. The identified fractional-order VSI-IM model is helpful to provide more insight into the dynamic behavior of the IM drive in the speed controller design (sensorless and sensored) and in predictive controls (model-based and model-free). The fractional VSI-IM models are identified at low, medium, and high frequencies. The best fit of these models is compared to the integer-order identified VSI-IM models. The experimental study clearly shows that the fractional-order VSI-IM identified model is superior to the integer-order VSI-IM model in terms of best fit with the actual hardware speed response.