遺伝的アルゴリズムによるエネルギープラントの機器構成設計法に関する研究

概要

論文の詳細を見る
An optimization method for the planning problem of energy plant configurations is proposed based on genetic algorithms. In such a planning problem, a plant configuration, i.e., types, models, and numbers of equipment, is determined so as to satisfy required energy damand conditions and to minimize the sum of plant facility cost and input energy cost. This is a combinatorial optimization problem similar to the knapsack problem, which is hard to find an optimal configuration due to the huge number of alternatives. In this paper, we apply a genetic algorithm to such an optimal planning problem by representing a plant configuration with bit vectors and by arranging cost functions so as to keep search performance superior against the deceptive problem. Its optimization performance is demonstrated by computational examples of a cogeneration plant and comparison results with a random search and a simulated annealing method.
一般社団法人日本機械学会の論文
1998-01-25