Research On Multi-agent Modeling And Simulation Method Of Train Control System Operation Scenario Based On Prometheus

The CTCS-3 train control system is the operation control technology equipment for high-speed railway over 300km/h,which plays an extremely important role in ensuring the operation safety and improving the running efficiency under the condition of high speed and high density tracking.It is a typical distributed real-time security demanding system.Once the train control system has a dangerous failure,the traffic interruption will affect the traffic efficiency and the serious traffic accident will occur.The CTCS-3 train control system is a product of deep integration of computer,modern communication,automatic control and other technologies,which is highly complex in system architecture,functional logic,control scenarios,interface relations and other aspects.Under different operation scenarios,CTCS-3 train control system has different interaction subjects,dynamic interaction behaviors,and complex control logic.Under complex operation scenarios,the potential risk factors and the superposition of risk factors of the system have strong diversity,concealment and complexity.In order to ensure the functional safety of the train control system,the risk analysis of the system is needed.However,the existing risk analysis methods mainly rely on experts’ experience,so it is difficult to analyze the risk in complex scenarios,and the correctness of the analysis results is difficult to verify.In view of the above problems,the paper considers the characteristics of the complex system such as distributed,concurrent,interactive,real-time,and hybrid of the train control system,using multi-agent modeling and simulation technology,studies the multi-agent modeling and simulation platform construction method for the train control system operation scenario.It provides simulation and verification technology for risk analysis in complex operation scenarios and is of great significance for ensuring traffic safety.Main research contents include:Firstly,based on the multi-agent modeling theory and the technical specification of the train control system,the multi-agent modeling method for the operation scenarios of the train control system was studied by using Prometheus method and PDT modeling tool.This method constructs the multi-agent model of the train control system from four steps:analysis of specific operation scenarios,system specification model,framework design model and Agent detailed design model.Then,based on the constructed multi-agent model and MASON simulation platform,the method of constructing multi-agent simulation platform for train control system operation scenarios is studied.In this method,the simulation platform is divided into five parts: system top-level module,data reading and processing module,visualization module,agent module and interactive packet data structure module.Finally,taking the RBC switching scene as an example,based on the line data,vehicle parameters and train control engineering data of Wuhan-Guangzhou Passenger Dedicated Line,the multi-agent model of RBC switching scene was constructed,and the simulation platform of RBC switching scene multi-agent was established.Based on the multi-agent model and HAZOP analysis method,the risk analysis of RBC switching scene was carried out,and the risk analysis data was injected into the simulation platform.The correctness of risk analysis is verified by simulation,and the effectiveness of multi-agent simulation method and platform is also verified.