High fitness population (HFP) with GA solution for solving unit commitment

Abstract

This paper presents an improved solution to optimal unit commitment (UC) by seeding best initial high fitness population (HFP) near or equal to global optimum solution to Genetic algorithm (GA). To direct the limited minimization option left in HFP in better way, easy GA mutation scheme is proposed that produces constrained satisfied populations, handle typical spinning reserve/time constraints and increases diversity option for GA to work near global optimum. The proposed algorithm performance is verified for systems of one-day scheduling period for 10–100 generating units. The test results reveal solution very near or close to optimum value achieved in initial population before the GA iteration starts. Result demonstrate the superiority of proposed scheme in term of number of iteration, cost and computation time then any other conventional methods / other computing techniques.