摘要
提出采用灰熵并行分析法引导粒子群算法求解多目标优化问题。灰熵并行分析法综合灰色关联分析法与信息熵的特点,对数据序列计算灰关联系数,同时并行地对数据序列计算信息熵及熵值权重,将灰关联系数与熵值权重结合求得灰熵并行关联度。通过粒子群算法对优化问题的多个目标构建与粒子数相同数量的目标值序列,计算每个序列的灰熵并行关联度值,利用该值作为算法适应度值的分配策略引导粒子进化。以10个典型作业车间调度问题为例进行实验,结果与差分进化算法及遗传算法进行比较,表明灰熵并行分析法可以有效引导各算法进化,使算法在收敛性和分布均匀性方面表现良好,且粒子群算法的优化结果要好于其他两种算法的结果。
A particle swarm algorithm is proposed based on the gray entropy parallel analysis method to solve the multi-objective optimization problems.The gray entropy parallel analysis method combines the characteristics of the grey correlation analysis method and information entropy.The grey correlation coefficient of the data sequence is calculated,meanwhile,the information entropy and the entropy weight are also calculated, then the grey entropy parallel correlation degree is got by combining the grey relational coefficient with the entropy weight.The objective value sequence of the multi-objective optimization problem is established by the particle swarm algorithm and the number of the objective value sequence equals to the number of particles in the algorithm.The value of grey entropy parallel correlation degree of each sequence is calculated and used as the distribution strategy of the fitness value to guide the particle evolution.Ten typical job shop scheduling problems are tested by the proposed method,and the results are compared with results gained by the differential evolution algorithm and the genetic algorithm.The experimental results show that the grey entropy parallel analysis method can guide the algorithm evolution effectively with good convergence and distribution performance,and the optimization results of particle swarm algorithm are better than those of the other two algorithms.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2014年第11期2233-2238,共6页
Systems Engineering and Electronics
基金
福建省自然科学基金(2014J01183)
福建省教育厅科技项目(JK2013006)
福州市科技计划项目(2012-G-131)资助课题
关键词
灰熵并行分析法
粒子群算法
多目标优化
作业车间调度
grey entropy parallel analysis method~ particle swarm optimization （PSO）
multi-objective optimization
job shop scheduling
