A New Approach for Solving Unit Commitment Problem Using Improved Shuffled Frog Leaping Algorithm

Abstract

ABSTRACT newlineThe main objective of this research work is to develop an evolutionary newlineapproach for the solution of Unit Commitment problem (UCP). Operating under the newlinepresent competitive environment, UCP is important in the Power System (PS) since a newlinesignificant amount of savings can be obtained from a sound Unit Commitment (UC) newlinedecision. newlineThe PS operates under continuous variations in load/demand. There is a wide newlinevariation in demand between weekdays and weekends and also between the peak and newlineoff peak hours of a day. Hence it is not economical to keep all the units on-line all newlinethe time. The operator should have the idea about the load forecast, economical newlineimplication and transition characteristics of generators while deciding the status of newlinethe units for a particular instant of the scheduling horizon. The main objective of the newlineUCP is to determine the minimum cost schedule of generators. The operating cost newlinemainly includes the fuel cost, start up cost and shut down cost. While scheduling the newlinegenerators economically, the system demand and reserve requirements impose the newlineglobal constraint. Also the individual units impose local constraints depending on the newlinefuel, capacity and characteristics of each unit such as capacity, ramp rate, minimum newlineup/down time, valve point effect, emission etc. newlineDetermination of minimal cost generator status is very difficult since it newlineinvolves many constraints. The UCP is a complex, non linear, mixed integer newlineoptimization problem. The optimal solution of UCP can be obtained by simultaneous newlinesolution of two sub problems. (i) Solution of mixed integer problem of determining newlinethe ON/OFF status of the units over the scheduling horizon considering the system newlinepower and reserve requirements. (ii) Solution of quadratic programming problem of newlineoptimal dispatch among the committed units considering the unit constraints. The newlinecomputational complexity of UCP increases with number of generators (Ng), newlinescheduling period (T) and no of constraints. Complete enumeration of all possible newlinesolutions can give the optimal schedule