Automatic generation method of virtual subway passenger flow big data based on AlphaZero thought

The invention discloses an AlphaZero thought-based automatic generation method of virtual subway passenger flow big data. The method comprises the steps of S1, performing initial condition setting according to the condition of a subway operation line and expert experience; s2, generating virtual subway passenger flow time distribution rule and space distribution rule data according to initial condition setting; and S3, generating virtual subway passenger flow microscopic big data according to the generated time distribution rule and space distribution rule. According to an AlphaZero thought, virtual passenger flow big data is generated to replace real data which is difficult to obtain and low in representativeness so as to solve related problems of a subway train dispatching scheme. 本发明公开了一种基于AlphaZero思想的虚拟地铁客流大数据的自动生成方法，包括：S1：根据地铁运行线路的情况和专家经验进行初始条件设定；S2：根据初始条件设定生成虚拟地铁客流时间分布规律与空间分布规律数据；S3：根据生成的时间分布规律和空间分布规律生成虚拟地铁客流微观大数据。本发明依据AlphaZero的思想，生成虚拟客流大数据代替难获取且代表性不强的真实数据，以用于地铁列车调度方案的相关问题。