Robust speed control design for electric motors using the Kharitonov theorem and genetic algorithms: an experimental study

Abstract

Electric motor drives are subjected to severe oscillations that might cause motor shaft fatigue and consequent breakdown during system operation. In this paper, a design of robust proportional integral (PI) controller for induction motor (IM) drives is presented. Uncertainties due to different load conditions of the drive system are modelled by a transfer function with interval coefficients. Instead of stabilising an infinite number of polynomials, the Kharitonov theorem is used to design a robust controller by simultaneous stabilisation of only four polynomials. The controller parameters are optimised so that the maximum eigenvalue among the four polynomials is pushed to the left in the complex plane. This is accomplished using genetic algorithm (GA) optimisation to achieve the maximum possible rapidity of the system. The proposed controller is tested experimentally using the hardware-in-the-loop (HIL) technique. The simulation and experimental results show the validity and effectiveness of the proposed controller as compared to the auto-tuned PID control at different loading conditions.