论文部分内容阅读
遗传算法被广泛应用于求解车间作业调度问题(JSP),但遗传算法具有最优参数难以确定的问题。对此,该文提出了一种基于神经元动态规划(NDP)的遗传算法NDP-GA。该文将遗传算法用M arkov决策过程模型描述,建立了M arkov决策过程最优策略与遗传算法最优参数之间的联系。在此基础上,用神经元动态规划逼近M arkov决策过程的最优策略,并用学习到的策略指导遗传算法最优参数的选择。数值计算结果表明,该文提出的算法能自动收敛到最优遗传参数,并在求解JSP问题时能稳定地得到满意解。
Genetic algorithm is widely used to solve job shop scheduling problem (JSP), but the genetic algorithm has the problem that the optimal parameters are difficult to determine. In this paper, a new genetic algorithm NDP-GA based on neuron dynamic programming (NDP) is proposed. In this paper, the genetic algorithm is described by M arkov decision process model, and the relationship between the optimal strategy of M arkov decision process and the optimal parameters of genetic algorithm is established. Based on this, the optimal strategy of M arkov decision-making process is approximated by neuron dynamic programming, and the learning strategy is used to guide the selection of optimal parameters of genetic algorithm. The numerical results show that the proposed algorithm can automatically converge to the optimal genetic parameters and can obtain satisfactory solutions stably when solving JSP problems.