Non-convex economic dispatch: A direct search approach

With increasing fuel prices and restructuring of the power industry, the non-convex economic dispatch (NED) problem may become a more important problem due to the inclusion of non-smooth cost functions. It is widely recognized that a proper schedule of available generating units may save utilities millions of dollars per year in production costs. This paper develops an enhanced direct search algorithm (EDSA) to solve the NED problem including transmission losses. The EDSA incorporates the parallel nature of evaluation programming into the direct search method to enhance its search capacity that leads to a higher probability of obtaining the global optimal solution. Many nonlinear characteristics of the generator, such as prohibited operating zones and non-convex unit cost functions can be handled properly in the direct search procedure. Several heuristic strategies are also used to improve the EDSA solution process. Numerical experiments are included to demonstrate the proposed approach can obtain a higher quality solution with better performance.