双种群变异策略改进RWCE算法优化换热网络

Optimization of Heat Exchanger Network Based on Improved RWCE Algorithm with Double Population Mutation Strategy

鉴于强制进化随机游走算法概率接受差解策略的两面性,一方面会出现差解代替仍有进化潜力的解,打断个体可能存在的进化路径,另一方面进化后期个体变异能力仍不够强,难以跳出局部最优值,提出了一种双种群变异策略。在基础种群之外,设置了2个特殊种群,周期性地接受基础种群中对应个体的当前最优解,并分别对特殊种群中个体进行接受差解概率的差异化处理,在周期结束回赋基础种群个体的最优值。通过算例验证,将改进后算法应用于换热网络优化,取得了较好的结果。

In view of the two sides of the probability acceptance of the worse solution in the forced evolution-ary random walk algorithm, on the one hand, the worse solution will replace the solution which still has the evolutionary potential and interrupt the individual’s possible evolutionary path. On the oth-er hand, the individual’s mutation ability is not strong enough to jump out of the local optimal value in the late evolution stage, so a double population mutation strategy is proposed. In addition to the basic population, two special populations are set up to accept the current optimal solution of the corresponding individuals of the basic population periodically, and the individual in the special population is treated differently to accept the worse solution probability, and the optimal value of the individual of the basic population is given back at the end of the cycle. The improved algorithm is applied to the optimization of heat exchanger network and good results are obtained.