【摘要】 城市智能交通体系属于空间信息系统,致力于提高交通效率、改善交通安全等。在已知个体行为交通序列的前提下,一项有意义的工作是观察个体如何有效地统筹安排其日常生活中行为的时间规划和路径选择,这将有助于对交通状况和需求的预测。该文在增强学习的框架下对上述的行为交通序列进行了优化和仿真。实验结果表明,增强学习不仅可以获得最佳的策略,而且在个人由于一些不可预测的事件而被迫偏离原来的最优安排时,仍可以依据学习过程中积累的信息明智地逐步调整过来。更多还原

【Abstract】 The research and development of intelligent transportation system(ITS) is prosperous in recent years.When the activity-travel sequence has been specified, it is meaningful to observe how the person arranges time and locations for these activities reasonably,which will benefit the prediction of transportation status and demand.This paper optimizes and simulates the activity-travel sequence under the framework of reinforcement learning.The experiments reveal that reinforcement learning can not only find the best strategy,but also make the person be able to react wisely when he is thrown off his optimal arrangement because of unforeseeable events.更多还原