An improved evolutionary programming approach to economic dispatch

This paper presents an improved evolutionary programming (EP) approach to power system economic dispatch. In the improved EP approach, quasi-random sequence (QRS) instead of pseudo-random sequence (PRS), is employed to generate much more evenly distributed initial population to avoid causing rapid clustering around an arbitrary local optimum. Furthermore, the population diversity is introduced to update the mutation range adaptively in the evolutionary process. The effectiveness of the proposed approach is demonstrated through its evaluation on two economic dispatch problems.