Differential Evolution Technique with Random Localization for Tuned Reactive Power Dispatch Problem

Abstract

This article explores the capability of a differential evolution technique applied with a concept of random localization for solving the reactive power dispatch problem. Being a strong optimization technique, differential evolution is still suffering from the problem of slow convergence at the global minima region. Reactive power dispatch, on the other hand, is an active power loss minimization problem where bus voltages are to be kept within a system-stable margin. In this article, modified differential evolution is chosen as an optimization tool for studying few parameters to solve the reactive power dispatch problem in case of IEEE 14- and 30-bus systems. The applied modified differential evolution technique shows much improved results compared to basic differential evolution technique. The obtained results are also compared with different soft-computing techniques, including differential evolution and another modified differential evolution to substantiate the effectiveness of the proposed approach.