磁悬浮列车跨系统运行Petri网模型

根据磁悬浮列车跨系统运行需求, 研究了其运行控制系统的总体框架, 明确了需要增加的功能子系统。基于系统理论, 采用Petri网对系统关键属性、列车运行过程及各子系统的功能进行了层次化的建模。最高层模型描述系统整体关键属性, 低层模型描述列车运行过程及可靠性。此模型可用来定量分析磁悬浮列车系统层面上跨系统运行时, 失败率与各子系统部件可靠性之间的关系。如每年磁悬浮列车跨系统运行失败次数不超过1次, 则连接相邻列控系统的2个通信网, 其失效率都需低于10^-6次·h^-1。当列车跨系统运行触发时间分别为0.2、2.0min, 步进时间分别为4、16min时, 则跨线运行失败率分别为1.95×10^-5、1.65×10^-5次·h^-1。仿真结果表明: 列车跨系统失败率随a网和b网可靠性的提高而降低, 同时随着跨系统触发时间和步进时间的增加而降低。层次化建模分析方法可以根据系统层面的关键属性要求, 定量确定各子系统部件的可靠性需求。

Petri net model of maglev train running across different control systems

The general framework of running control system on maglev train was studied according to the running requirements of maglev train running across different control systems.Functional subsystems need to be added was defined.The hierarchical models of system key attributes,maglev operation procedures and the subsystem function were built based on the system theory by using Petri net.The key attributes of whole system were described by the highest model,and the operation procedures of maglev train and the reliabilities of subsystems were presented in the lower level model.The relationship between the failure rates of maglev train running across different control systems and the reliabilities of subsystem components was quantitatively analyzed with the model.It is pointed that the loss ratio of network connecting neibouring control systems should be lower than 10-6 times per hour when the required failure number of maglev train running across different systems is no more than 1 time per year.The failure rates of maglev train running across different control systems are 1.95×10-5 and 1.65×10-5 times per hour when the triggering times equal 0.2 and 2.0 min respectively,and the stepping times equal 4 and 16 min respectively.Simulation result shows that the failure rates of train running across the boundary decrease when the reliabilities of a and b networks are improved,or the triggering time and stepping time of train are prolonged.The reliability requirements of subsystem components based on the required key attributes of system level are quantantatively identified by using the proposed approach.9 figs,14 refs.