Data-driven Highway Macro-micro Integrated Traffic Simulation And Control Optimization

A highway,as the artery of urban road network,can significantly enhance the efficiency of transportation services owning to its high speed limit and good traffic condition.However,the uneven travel demand exacerbate the problem of holiday traffic congestion on major arterials entering and exiting the city.This thesis develops a datadriven macro-micro integrated simulation and a hierarchical control model to address the problem of highway congestion.First,the macro-micro model parameters are calculated using actual data.Then,the macro model gives the optimal control parameters online depending on the current micro-road network traffic states.Control parameters are then transformed into the real control strategy via the microscopic model.In this thesis,two highway scenarios are chosen for applying the control strategy: a complete highway network and an urban road network with its downstream highway.Both scenarios are applied to verify the effectiveness of the proposed macro-micro integrated control model,which is to alleviate the highway congestion.The research contents of this thesis are given as follows:First,a data-driven macro-micro composite simulation model is established.The SUMO microscopic simulation software is used to realize it,which is based on an actual highway’s road network and detector data.The driving parameters of different vehicle types in each road section are tested by optimizing the speed and flow of the road section through splitting the road section and vehicle type;the macroscopic traffic flow model is based on the microscopic simulation model’s simulation data.The macroscopic fundamental diagram is used to build the macroscopic traffic flow model of each road segment,and the highway traffic flow conservation equation is formulated,which is utilized to represent the evolution of highway traffic state and the basic law of traffic flow.Second,a hierarchical control strategy is proposed for a highway road with its upstream urban road network.This control strategy is from an overall perspective.At the macroscopic level,the MPC model determines the current state parameters obtained from the road section of the micro-simulation while the state parameters of the road section at multiple time in the future are obtained to minimize the actual number of vehicles on the road network.The objective is the difference between the actual number of vehicles and the optimal one.The genetic algorithm is adopted to determine the optimal control parameters at the next time as the input of the micro-model.In this way,the number of vehicles in the road network can always be maintained around the optimal number of vehicles to ensure the road network’s overall service capabilities.Finally,the simulation experiment is conducted by using the SUMO simulation platform.The control parameters derived by the macroscopic model are sent to the SUMO micro model via the Tra CI interface during secondary development of the SUMO simulation software to acquire the associated indicators.Also,by using the big data platform,the visualization of highway detection data and the road traffic flow information before and after control is provided in order to construct the highway data display interface.