论文部分内容阅读
针对订单生产型企业在订单接受决策过程中的不确定性,基于强化学习的思想,在考虑生产成本、延迟惩罚成本以及拒绝成本的前提下,引入顾客等级这一要素,从收益管理的角度建立了基于半马尔可夫决策过程的订单接受模型.在此基础上,提出了基于SMART算法的最优订单接受策略求解方法,旨在最大化订单生产型企业的长期利润.仿真实验结果表明:基于SMART算法得到的订单接受策略要优于基于先来先服务方法得到的订单接受策略;同时,针对考虑顾客等级的仿真实验及数据分析结果,也验证了引入顾客等级这一要素的必要性和重要性.
In view of the uncertainty of order-oriented manufacturing enterprises in order acceptance decision-making process, based on the idea of reinforcement learning, the introduction of customer grade is based on the concept of revenue management, considering the cost of production, the cost of delaying punishment and the cost of rejection. Order acceptance model based on semi-Markov decision process.On this basis, the optimal order acceptance strategy based on SMART algorithm is proposed to solve the long-term profit of order-producing enterprises.The simulation results show that: SMART algorithm is superior to the order acceptance strategy based on the first-come-first-serve method. Simultaneously, the simulation experiment and data analysis considering the customer level also verify the necessity and importance of introducing customer grade Sex.