Using evolutionary computation to solve the economic load dispatch problem

The classical approach to the Economic Load Dispatch Problem (ELDP) seeks to minimize the cost of generation subject to the usual constraints. If the transmission losses are also to be taken care of, a common method (lambda - iteration procedure) involves adding the cost of transmission losses charged at incremental cost of received power to the cost of generation. This combined cost function forms the objective function to be minimized. However it is desirable that the transmission losses be dealt with separately in the minimization process, without computing the incremental cost of received power. Genetic Algorithms (GAs) offer a suitable and robust approach to meet the twin objectives of cost minimization and loss minimization simultaneously. This paper presents the development of the GA for solving The ELDP. It also suggests a convenient technique for representing the search space. Finally, the results of the algorithm are analyzed by comparing them with those obtained from a direct random search algorithm, called the LJ algorithm.