Research On Dynamic Traffic Estimation And Perimeter Robust Optimal Control Of Urban Road Region

As China’s new urbanization reform continues to progress,more and more workers are gathering in cities,and the gathering of population and industry has brought "urban diseases" while injecting vitality into urban development.The mass gathering of population makes the demand and supply of urban transportation imbalanced.Traffic congestion is related to the regional traffic demand,and it will occur when the capacity of the urban region network reaches the maximum.Therefore,reasonable control of the input flow of the urban region network is the key to solve traffic congestion,regional perimeter control usually assumes sufficient knowledge of the accumulated state as well as inflow flows in the region,but it is difficult to measure the accumulation state and inflow demand of the region based on OD in practice,at the same time,there exists some noise in the measured values.In order to solve these problems,it is necessary to propose a reasonable estimation method to obtain the accurate traffic demand.For this purpose,this paper conducts a research on the dynamic estimation and perimeter robust optimal control of urban region network based on macroscopic fundamental diagram,which consists of the three main components as follows:(1)Dynamic estimation of urban region network based on extended kalman filter.First,on the basis of the multi-region macroscopic fundamental diagram model,an extended kalman filter-based flow estimation model is established,which is able to estimate dynamic flow of multi-region urban region network.Then,four measurement combinations are proposed.Finally,the effect of the proposed estimation model is verified using a numerical example based on the regional urban network in Yokohama,Japan.(2)Robust optimal perimeter control for urban regions;Firstly,a perimeter control model based on model predictive control is established on the basis of a multi-region macroscopic fundamental diagram model,then the robust counterpart of the control model based on the scenario set is carried out considering the uncertainty of the state matrix and control matrix.Secondly,the perimeter robust optimal control algorithm is designed,and simulation experiments are conducted using the Yokohama regional road network in Japan to verify the robustness and the control effect of the perimeter robust optimal control algorithm,respectively.(3)Robust optimal perimeter control method based on dynamic traffic estimation of traffic flow.The integrated algorithm of dynamic flow estimation and perimeter robust optimal control is set up by combining the estimation model and perimeter robust optimal control model.A case study is conducted on the urban region network in Wangjing area of Beijing to verify the estimation effect and control performance of the integrated algorithm.The results show that the method can improve the effect of noise elimination and has good effectiveness of filtering,at the same time,the accumulation of vehicles in every subzone is able to reach the desired state in the case study.Meanwhile,the effects of different combinations of measurement on the estimation performance and control performance are considered in the case study,we calculate the estimation and control effects of different combinations of measurement by introducing parameters respectively.The result show that the dynamic flow estimation of urban region and perimeter robust optimal control strategy has wide applicability.