Optimization And Case Study Of Bus Route Design For Different Vehicle Types

With the acceleration of urban motorization,it has been an astounding growth in automobile ownership,leading to a series of traffic related problems such as urban congestion and environmental pollution.As a high-reliability,large-capacity,safe and economical mode of transportation,public transportation is highly valued by the government.Being regarded as an important part of the urban public transportation system,the efficiency of bus transit,which depends largely on stop locations and service headway,has a significant impact on user cost,operator cost,and pollutant cost.The optimization of bus stops locations and service headway can achieve to reduce total system cost,thus attracting more urban habitants to choose public transportation and then alleviating the traffic congestion and solving vehicle emission problem.In this thesis,we firstly propose a bi-directional(upstream and downstream)and muti-period continuum model(peak hour and off-peak hour)to minimize the total system cost.The decision variables contain upstream/downstream stop density functions and peak/off-peak service headway.In the proposed model,the environmental factor is considered.On the basis of Continuum Approximation(CA)approach,a discrete model is build in order to verify accuracy.Here includes the major research contents and conclusions as following:(1)This paper proposes a bus route design model based on CA,with generalizedsystem cost function including passenger's travel cost,operator cost,andpollutant cost.Upstream/ downstream stop density functions andpeak/off-peak hour headways are chosen as decision variables.Analyticalsolution of optimal stop density functions and service headways areillustrated by using Lagrange Multiplier method while satisfyingKarush-Kuhn-Tucker(KKT)conditions.(2)The model is implemented in the real-world bus route 7 in Yaan City(China).By employing the on and off data into proposed model,the bus route designis optimized when four different vehicle types are operating separately on the route.The current system cost items are compared with the optimized system cost items.Then,the discrete model which is based on continuum model is developed.The comparison between the continuum model system cost and discrete model system cost is made.The results indicate that:(1)The optimization leads to huge cut in total system costs(by up to 60%)for four kinds of bus systems which verify the adaptability of proposed model in transit route design.(2)The outcomes of discrete model and continuum model are in the neighborhood,with error less than 4%,which verify the accuracy of CA model in real transit route design.On the basis of the proposed CA model,different driving strategies are considered into optimization.The generalized cost could be influenced when buses are under different driving regimes(accelerating,cruising,coasting,and decelerating).There are four kinds of driving strategies while buses are operating between stops,that is: accelerating-decelerating;accelerating-cruising-decelerating;accelerating-coasting-decelerating;and accelerating-cruising-coasting-decelerating;The upstream/downstream stop density functions,peak/off-peak service headway,and upstream/downstream inter-stop coasting speed are chosen as decision variables.Based on the proposed CA model,a discrete model is constructed in order to verify accuracy.In the end,the life-span cost effectiveness comparison is conducted amongst four different vehicle types.Here includes the major research contents and conclusions as following:(1)The optimal system cost decreases almost 50% comparing withcurrent system cost,with average stop spacing increased by up to30%.The difference between the discrete model system cost andthat of CA model is less than 3%,verifying the adaptability andaccuracy of CA model.(2)The optimization of bus route design considering driving regimesand environmental factor indicates that introducing coasting regimeinto optimization will increase the passenger cost(by 14%),whiledecrease operator cost(by 20%).The analysis of the effect ofcoasting regime suggests that,introducing coasting regime intooptimization is not profitable for new-energy buses,but forconventional buses,it can help to reduce energy consumption whichis beneficial.(3)Considering the increase of per-km maintenance cost ofconventional buses(CNG buses and diesel buses)with year,the costeffectiveness comparison is conducted amongst four vehicle types ina life span(normally 8 years).The results indicate that,consideringthe pollutants costs for conventional buses(i.e.,CNG bus and dieselbus),supercharge bus will outperform other bus types after 8 years,which is still out of a life span.Thus,clean-energy buses are lesscompetitive than conventional buses unless subsidies are provided.