A new control technique for active power filters using a combined genetic algorithm/conventional analysis

In this paper, the computational problems associated with the optimization techniques used to evaluate the switching patterns for controlling variable-characteristics active power filters are presented and critically analyzed. Genetic algorithms (GAs) are introduced in this paper to generate a fast and accurate initial starting point in the highly nonlinear optimization space of mathematical optimization techniques. GAs tend to speed up the initialization process by a factor of 13. A combined GA/conventional technique is also proposed and implemented to reduce the associated computational burden associated with the control and, consequently, increasing the speed of response of this class of active filters. Comparisons of these techniques are discussed and presented in conjunction with simulation and practical results for the filter operation.