Study On Lane Reservation Problems From The Perspective Of Transportation Network

Nowadays,transportation has become an indispensable part in modern life.However,heavy traffic congestion due to high urbanization and rapid increase of vehicles has caused low transportation effiency and huge amounts of urban pollution.In this context,to meet special transportation requirements and improve the performance of transportation systems,lane reservation strategy,as a flexible and economic traffic management strategy,has been widely implemented in real life.The majority of studies about lane reservation in the literature focus on investigating the impact at a microscope level,e.g.,a single link or corridor.However,they cannot guide optimal lane reservation decision at a macroscopic network level.Lane reservation strategy occupies existing road resources and will cause negative impact on the normal traffic.Therefore,appropriately reserving lanes from the macroscopic network view of point plays an important role in reducing their negative impact.This thesis investigates lane reservation problem from the perspective of transportation network.The studied problems consist of optimally selecting lanes from an existing network to be reserved to meet special transportation needs or improve bus transportation performance with the objective of minimizing the negative impact caused by lane reservation.The main contribution brought by this thesis is summarized as follows:(1)This thesis first investigates large-size automated truck lane reservation problem.An improved integer linear program is developed.Its several special cases are identified to be classical combinatorial optimization problems.Based on the derived properties,a two-phase exact algorithm is developed.Computational results on benchmark and newly generated larger-size instances indicated that the proposed algorithm signicantly outperforms the existing method.(2)Secondly,a robust lane reservation problem for meeting special transportation needs arising from large-scale events is studied.For this problem,a multi-objective integer linear program is formulated.Its preprocessing technique is explored.Then,an ?-constraint and cut-and-solve combined method is developed to generate the Pareto front.Conputational results on a benchmark instance and randomly generated instances show that the proposed algorithm is more efficient than the integer programming method based on the commercial optimization software CPLEX.(3)Thirdly,this thesis addresses a bus lane reservation problem for rapid and reliable bus transit.Two integer linear programs are developed for this problem,and valid inequalities are explored to reduce the search space.An improved cut-and-solve algorhtm based on an improved pirecing cut is proposed to obtain optimal solutions.Then,to more efficiently solve the problem,especially for large-size problems,a kernel search based heuristc is developed to obtain high-quality solutions within shorter time.Extensive numerical experimental results validate the performance of the developed models and approaches.(4)Finally,a bus lane reservation and bus line design integrated optimization problem is investigated.For this problem,a nonlinear multi-objective mixed integer program is first formulated.Then,the nonlinear model is transformed into an equivalent mixed-integer linear program,and several valid inequalities are proposed to reduce the search space for Pareto optimal solutions,and the problem complexity is analyzed.Finally,an exact ?-constraint method is proposed to obtain the Pareto front.Computational results on a benchmark instance and randomly generated instances show that the proposed method can generate the Pareto front within reasonable time and the obtained lane reservation scheme can greatly reduce bus travel time at relatively low negative impact.