基于免疫遗忘的车间调度算法

An Immune Forgetting Algorithm for Job Shop Scheduling Problems

为了解决车间调度NP组合优化的难题,提出了基于免疫遗忘的车间调度算法。算法在基于克隆选择方法能同时搜索解空间的不同区域以及能保持种群的多样性的功能的基础之上,又加入了遗忘单元,形成了一个来自于抗体群中较好抗体组成的种群,从而可以实现在每次迭代中对遗忘单元进行邻域搜索。算法使抗体群和遗忘单元共同进化,并互相影响,使算法在寻找满意解上得到优化。仿真实验表明,算法能找到比遗传算法更好的满意解。

In order to solve NP- shop scheduling combinatorial optimization problems, an immune forgotten algorithm for job shop scheduling is proposed. Based on the clonal selection method which can search the solution space of the different regions and maintain the population diversity of the functions, the algorithm has joined the forgotten u- nit, forming an antibody from a base population consisting of a good antibody, which can be achieved in each itera- tion of the forgotten unit neighborhood search. The algorithm makes antibodies of cell clusters and forgotten units co - evolution and interact. The algorithm can be optimized in finding the satisfactory solution. Simulation results show that the algorithm can find better satisfactory solution than genetic algorithms does.