A Study On Park And Ride System Analysis, Evaluation And Optimization

Park and ride (P&R) facilities encourage commuters to enter the city's central areas by transferring from single/low occupancy vehicles (private cars) to high occupancy vehicles (public transport). Successful experiences with P&R facilities in European and American cities have concluded the P&R is a viable and promising travel demand management strategy, and also an effective transport mode to support modern city's development.Original P&R facilities have traditionally been designed as components of urban public transit, for the purpose of primarily expanding transit's market. Not a large amount of customers can be attracted by those P&R facilities. The role of the P&R facility is rapidly changing. Modern P&R system with core of P&R activity conducts the demand estimation, facility siting and operation management, and will takes the majority of commuter trips, plays more important role in urban transport.Successful P&R planning can not be separated from good understanding and deep analysis of P&R behavior. Traditionally, P&R behavior research approach is statistical analysis on factors affecting individual choice decisions based on data from questionnaire survey. Abundant experiential conclusions about P&R behavior characteristics have been achieved as good guidance for P&R planning and design in practice. However, there are frequent interactions among various travel behaviors on urban transport networks, which can not be ignored to ensure logical results especially when transport supply and demand conflict more seriously, and travel alternatives become diversiform today.Network equilibrium analysis is a powerful tool for covering network performance, such as congestion effects, capacity limits, which has been widely applied in research fields of urban transportation analysis and optimization for decades. Up to now, few attentions have been paid on P&R network equilibrium study. This paper is trying to conduct a comprehensive research on P&R behavior analysis and system optimization. The work is an expenditure of transportation theory, and provides some insights and guidance for P&R planning and management.Firstly, a network equilibrium formulation was proposed for modeling the P&R services in a stochastic transportation network with capacity constraints. In the propose model, it is assumed that commuters can complete their journeys by three options:using auto mode, walk-metro mode or the auto-metro mode (referred to as the P&R mode). The proposed model simultaneously considers commuters'travel choices on travel mode, route/path and transfer point, as well as their parking choice behavior nearby the destination. The proposed model can be formulated as a fixed-point problem. An augmented Lagrangian dual algorithm embedded by the method of successive averages (MSA) is developed to solve the proposed model. Numerical results show that adding a P&R facility into the network might lead to a paradoxical phenomenon, i.e., total network travel costs after adding a new P&R facility become larger than before instead.It follows with an investigation on the effects of ATIS on P&R behavior in a multimodal transportation network. A multi-class probit-based stochastic network equilibrium formulation was presented. It can be formulated as a VI problem and solved by a simulation-based heuristic. The commuters, which are classified into two types of equipped and unequipped with ATIS, would both make their travel choices following a probit-based stochastic manner. Numerical results show that the RGS and PIS have contrary impacts on P&R demand, and the improvement of ATIS quality will draw more commuters to auto mode.Secondly, P&R reliability and mode reliability are introduced as new performance indices for the evaluation of P&R network uncertainty. Travelers' disutility is formulated as sum of certain cost and uncertain cost to explicitly describe effects of travel time uncertainty on individuals'decisions making, such as mode choice, route choice and parking facility choice. A VI (variational inequality) model is established under SUE condition. A heuristic solution algorithm that uses a combination of the Monte Carlo simulation approach with the method of successive averages is designed to solve the proposed VI model. Sensitivity analyses for effects of the capacity of parking facilities, dispatching frequency of the metro line and metro fare on network reliability and travelers'behavior were was carried out by numerical experiments.Finally, three types of P&R system optimization problems are studied, involving P&R facility siting, P&R network design and P&R pricing. A bi-level programming model is proposed based on stochastic user equilibrium (SUE) conditions, where the upper level aims to maximizing the total social welfare by determining to sit P&R facilities, while the lower level is a network equilibrium model considering travelers'behaviors. A heuristic algorithm which is a combination of genetic algorithm (GA) and method of successive average (MSA) is designed to solve the presented model. Numeric examples are given to test the effectivity of the proposed model and algorithm, which are also applied to sit P&R facilities on Chengdu city as a demonstration of the ability of solving siting problem on large scale networks.Then, this paper studies the metro line design problem in a P&R network with elastic demand. The interaction between the traffic planner and manger and the network users can be also formulated as a bi-level hierarchical problem. A branch-and-bound method is developed to solve the proposed model. A numerical example is used to illustrate the application of the proposed model and solution algorithm.The ending is investigation on the parking pricing problem of various parking facilities, i.e., off-street parking lots, street parking lots and P&R lots, under different market regimes, such as monopoly, oligopoly and social optimum. The interaction between parking operators and the travelers in the system is also described as a bi-level programming problem. The upper level is a decision model for optimizing the parking charging level, whereas the lower level is a traveler choice model. A sensitivity analysis based algorithm is adopted to solve the proposed model. Numerical experiments are used to provide insights of market regimes impacts on network performance and parking pricing structure.