改进灰狼优化算法求解模糊车间调度问题

模糊车间调度问题是复杂调度的经典体现，针对此问题设计优秀的调度方案能提高生产效率。目前对于模糊车间调度问题的研究主要集中在单目标上，因此提出一种改进的灰狼优化算法（improved grey wolf optimization，IGWO）求解以最小化模糊完成时间和最小化模糊机器总负载的双目标模糊柔性作业车间调度问题。该算法首先采用双层编码将IGWO离散化，设计一种基于HV贡献度的策略提高种群多样性；然后使用强化学习方法确定全局和局部的搜索参数，改进两种交叉算子协助个体在不同更新模式下的进化；接着使用两级变邻域和四种替换策略提高局部搜索能力；最后在多个测例上进行多组实验分析验证改进策略的有效性。在多数测例上，IGWO的性能要优于对比算法，具有良好的收敛性和分布性。

Improved grey wolf optimization algorithm for fuzzy Job-shop scheduling problem

Fuzzy Job-Shop scheduling problem is a classic embodiment of complex scheduling, and an excellent scheduling scheme designed for this problem can improve production efficiency. At present, the research on fuzzy Job-Shop scheduling problem mainly focused on single objective, this paper proposed an improved grey wolf optimization(IGWO) algorithm to solve the bi-objective fuzzy flexible Job-Shop scheduling problem to minimize the fuzzy completion time and the total load of fuzzy machines. This algorithm proposed a two-layer coding method to make IGWO discretization, designed a strategy based on HV contribution degree to improve population diversity. Then it used reinforcement learning method to determine global and local search parameters, improved two crossover operators to help individuals evolve in different update modes, and used two-level variable neighborhood and four replacement strategies to improve local search capability. Finally, this paper carried out several groups of experiments on several examples to verify the effectiveness of the improved strategy. In most test cases, the performance of IGWO algorithm is better than the comparison algorithm, with good convergence and distribution.