Scenario selection by genetic algorithm for evaluating power resource planning

Power resource planning which decides the construction time of generators is affected by several uncertain factors: load demand, fuel cost, and others. To evaluate power resource planning involving several uncertainties, planners have to prepare many scenarios and analyze whether the particular resource plan is feasible for every scenario. Of course, it takes a long time to analyze even one scenario. Therefore, in order to realize an efficient scenario analysis, effective scenarios must be selected, and the plan must be evaluated for these scenarios. In this paper, the authors propose a new algorithm for evaluating power resource planning with a genetic algorithm. Specifically, (1) a new preferable scenario selection algorithm, and (2) a new multi-objective scenario analyse algorithm are developed. Several numerical results for a new scenario selection algorithm are presented to demonstrate the effectiveness of the proposed algorithm.