The Research Based On Complex Event Processing Method For Transportation Internet Of Things Simulation System

Proactive complex event processing means predicting possible states of system and performing some operations to avoid the occurrence of the state according to a large number of uncertain events.Taking some measures to eliminate or mitigate the threat or using the chance of prediction can significantly improve the quality of people’s lives and prevent environmental pollution and reduce economic losses.The proactive complex event processing method that applied to the traffic simulation system can forecast traffic state,and take corresponding measures to avoid traffic congestion and reduce carbon dioxide emissions which can optimize public transportation and improve the efficiency of walker.Therefore building a traffic network simulation system to support active complex event processing has important research significance and practical significance.Existing traffic simulation system is not combined with Agent technology and lack of the ability of learning and decision-making.When performing timing study of traffic lights,the effect is not good in the aspect of vehicle simulation.So this paper propose to build an intelligent traffic simulation system based on microscopic traffic simulation tools SUMO and Java Agent Development Framework JADE.In this system,we build a vehicle and driver Agent model for every car and driver of traffic.In order to make the model more realistic,we use weighted linear by particle swarm optimization(PSO)algorithm to calibrate model parameter.This paper takes microscopic model parameter optimization of the vehicle and the macro decision-making of vehicle routing choice based on complex event processing method as the research focus,the main work is as follows.Firstly,a simulation framework of active traffic internet of things based on proactive complex event processing is proposed and realized.We use microscopic traffic simulation tools SUMO and intelligent Agent framework JADE to build an intelligent transportation network simulation system oriented in active complex event processing.This thesis uses a small amount of configuration to generate the actual configuration files,and realizes the traffic flow configuration based on all kinds of probability distribution to make it closer to reality.This thesis regards vehicles and drivers as a research unit,and then build the Vehicle-Driver Agent.In order to make the simulation more real,and simulate the characteristics of intelligent operation of the driver and the vehicle’s dynamic performance,we build micro modeling for the Agent and use the linear weighting particle swarm optimization to optimize the model parameters.In order to make the simulation more reality and support the active control better,this paper proposes a kind of Markov decision method which has the function of predicting.What’s more,it can perform dynamic allocation and macro-control for traffic route choice of simulation,and thereby easing congestion and improve traffic,improve the traffic efficiency.