Improved Directional Bat Algorithm Based Electric Power Dispatch

No single metaheuristic search algorithm can be adjudged universally best general-purpose optimizer. The performance of search algorithms mainly depends upon the weightage assigned to global and local search strategies. This paper proposed an improved directional bat optimizer to minimize the operating cost of the electric power dispatch (EPD) problem that establishes a balance between global and local search strategies. Improved directional bat algorithm exploits directional echolocation bat behavior, directional exploration, neighborhood search and opposition based learning for generation jumping. The directional bat algorithm acts as a global search tool whereas exploration in each direction and neighborhood search performs local search. Opposition learning improves convergence with diversity. An effect of valve-point loading introduces a discontinuity in cost characteristics. The EPD problem addresses energy balance, generator capacity, ramp-rate limits and prohibited operating zones (POZ) avoidance constraints. An iterative technique handles energy balance constraint. The generation is adjusted to avoid the violation of generation capacity, ramp-rate limit and POZ constraints. The proposed algorithm is verified on various electric power systems. The results verify that the proposed algorithm is a potential algorithm to solve EPD problems as it competes with recent existing algorithms undertaken for comparison