一种粒子群和改进自适应差分进化混合算法及在生产调度中的应用

差分进化算法是一类基于种群的启发式全局搜索技术,但传统的差分进化算法存在停滞现象,容易使算法收敛停止。虽然之后出现了各种版本的自适应差分进化算法,但没有考虑到当代个体的适应值是否向着最优个体的适应值逼近,因此本文提出了一种新型的自适应差分进化算法FMDE。考虑到粒子群算法和差分进化算法类似,为了充分发挥两种算法的特点,提出了自适应差分进化和粒子群的混合算法PSO_FMDE；最后采用测试数据集对性能进行分析。实验结果表明,该算法根据进化过程中的搜索进度自适应地确定变异率,使算法易于跳出局部最优解,以提高全局搜索能力。PSO_FMDE算法较单一算法而言,性能更优,更易于靠近全局最优解。

A hybrid algorithm based on the combination of different evolution and particle swarm optimization and Its Application on Flow Shop Scheduling Problem

Differential evolution algorithm is a heuristic global optimization technique based on population. But conventional differential evolution algorithm has a problem of stagnation that can stop the algorithm convergence. Although various versions of adaptive differential evolution algorithms have emerged, it does not take into account whether the fitness of the contemporary individual approximates to the fitness of the optimal individual. An adaptive differential evolution algorithm FMDE is proposed. In order to take advantage of different algorithms, a hybrid optimization algorithm is proposed, based on the combination of different evolution and particle swarm optimization. Finally, the performance is tested and analyzed by using the data of instance. The results show that PSO_FMDE can determine mutation rate adaptively, which enhances the probability of obtaining the global optimum. Comparing with single algorithm, PSO_FMDE algorithm has better performance and is easier to get close to the global optimal solution.