论文部分内容阅读
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。
Urban Intelligent Transportation System is a spatial information system dedicated to improving traffic efficiency and improving traffic safety. Given the sequence of individual behavioral traffic, it is of interest to observe how individuals can effectively plan their time planning and routing of behavior in their daily life, which can be helpful in predicting traffic conditions and needs. In the framework of enhanced learning, the paper optimizes and simulates the above traffic sequence. The experimental results show that enhanced learning can not only obtain the best strategy, but also adjust intelligently and gradually according to the accumulated information in the learning process when individuals are forced to deviate from the original optimal arrangement due to some unpredictable events.