Generation mix planning using genetic algorithm

Generation mix planning (GMP) is one of the most important planning activities in electric utilities. Optimal long term GMP is to determine the least cost capacity addition schedule (i.e., the type, location and number of each candidate plant) that satisfies forecasted load demands within economic criteria over a planning horizon. In this paper, a modified genetic algorithm (MGA), which can overcome the aforementioned problems of the conventional SGA to some extents, is developed. The proposed MGA incorporates the following two main features: (1) an artificial initial population is devised, which also takes the random creation scheme of the conventional GA into account; and (2) a stochastic selection of reproduction candidates from a mating pool. The MGA approach is applied to IEEE 14-bus test system, one with 2 existing power plants, 3 types of candidate plants and a 20-year planning period.