Switching Angle Optimization Based Genetic Algorithms for Harmonic Reduction in Three-Phase PWM Strategy

Abstract

In variable speed drive (VSD), it is desirable to reduce the harmonic effects, which causes current distortion and torque pulsation, besides, the harmonic power losses is an additional power losses that is introduced in the motor due to the presence of harmonic voltages.
 However, the problem of the high total harmonic current distortion (THD) still exists specially at low and medium speeds by using sub-optimal pulse width modulation (PWM) strategy. In the past to generate optimized PWM, is done by defining a general PWM in terms of a set of switching angles. Which result in a set of nonlinear equations in terms of the unknown switching angles. These equations are nonlinear as well as transcendental in nature. There is no efficient method that can be applied to solve such equations. The practical method of solving these equations is a trial and error process. Taking all the factors into account, a numerical technique can be applied to solve these set of nonlinear equations, but with some limitations.
 To overcome these limitations, Genetic algorithms (GAs) serves to search for optimal switching angles setting. In addition, the (THD) will be reduced, this lead to obtain the optimal PWM waveform and to simplify the practical implementation, and then improving the performance of the system output.
 GAs were employed as a search and optimization engine. Normally the tuning of the switching angles is a trail and error problem.
 In this paper, GAs provides a much simpler approach to off-line tuning of PWM switching angles than the rather complicated non-genetic optimization algorithms