基于DEA混合算法的模糊车间作业计划问题的研究

Research of fuzzy job-shop scheduling problem based on hybrid DEA-GA algorithm

针对以最小化制造跨度为目标,具有模糊加工时间的车间作业计划问题,采用梯形模糊数来表征时间参数,并应用可能性理论,在此基础上构建车间作业计划问题目标函数。为了对模糊环境下的车间作业计划问题进行有效求解,给出了一种DEA-GA混合求解算法,混合算法采用了DNA进化算法的分裂、变异和水平选择算子,然后利用遗传算法的交叉算子实现个体之间的交互,避免早熟收敛。仿真实验表明,该算法高效可行,与GA等优化算法相比,具有更快的收敛速度。

This paper studied the job-shop scheduling problem which had fuzzy operation time and aimed at minimizing makespan. For this problem,it introduced trapezoidal fuzzy number to denote time parameters,on which the aim function was constructed. After that,proposed a hybrid DNA evolutionary algorithm integrating the mechanism of DNA evolutionary algorithm with genetic algorithm to get perfect scheduling scheme. To escape immature convergence,combined the crossover operator of genetic algorithm with the operators in DNA evolutionary algorithm,including the division,level selection,mutation operators. Moreover,simulation results show that this algorithm is feasible and effective. Compared with other optimization algorithms such as GA,it has rapid convergence ability.