The Application of Evolutionary Computation Technique for Short-Term Unit Commitment Problem

Scheduling the operation of thermal generating units involves the selection of units to be placed in the operation and the allocation of load among them. This decision must be taken so as to minimize the operating cost subject to various constraints. This paper presents the evolutionary computation (EC) technique for solving short-term unit commitment (UC) problems in order to minimized total operating cost. Initial population of parent solutions is generated at random to form the basis of subsequent generations. Each contending solutions are evolved through mutation, competition and selection. The proposed technique was tested on a reported UC problem previously addressed by some existing techniques such as dynamic programming (DP) and genetic algorithm (GA).