Authors : S.K. Bath , J.S. Dhillon and D.P. Kothari
Abstract: This stidy explores the use of genetic algorithm and Hooke-Jeeves search methods to search for the optimum active power generation schedule of thermal power systems, so as to obtain the best compromised solution, in the multi-objective framework. The multi-objective problem is formulated to minimize non-commensurable objectives viz. operating cost, NO emission and variance of active power generation with x explicit recognition of statistical uncertainties in the thermal power generation cost coefficients, emission coefficients, power demands and hence power generations and bus voltages, which are considered random variables. Inequality constraint to maintain security of transmission lines with respect to active power flow and equality constraint of active power balance are considered in the form of objective functions to be optimized. The objectives are quantified by defining their membership functions using fuzzy set theory The solution set of such formulated problems is non-inferior due to contradictions among the objectives taken. Active power generations are searched by genetic algorithm and Hooke-Jeeves search methods in the non-inferior domain. Among the generated set of non-inferior solutions of power generation schedules, system operator chooses the set that provides maximum satisfaction level of the most under achieved objective in terms of membership function and is termed as fitness function. The validity of the proposed methods is demonstrated and results are compared for an IEEE system comprising of 25-nodes, 35-lines and 5-generators.
S.K. Bath , J.S. Dhillon and D.P. Kothari , 2005. Stochastic Multi-objective Generation Dispatch by Search Methods . Asian Journal of Information Technology, 4: 823-831.