Introducing Park and Ride Location Model and solving it using Genetic Algorithms in the GIS Environment

Nowadays one of the major problems that public transit system planners are facing is lack of enough availability of the system for private vehicles. So providing reliable and efficient public transportation for this group is an effective way for overcoming car dependency. One of the alternatives that can reduce this problem is Park-and-ride that has been used since 1930s as a travel demand management measure.For the effective use of park-and-ride facilities, lots of elements should be considered. Park-and-ride facilities in the U.S. and Japan are being well utilized and one of the major factors in theses countries for the successful operations is the location of facilities.Given this, this paper is devoted to the development of a GIS based two-parts method for locating park and ride facilities based on a defined set of criteria, which will include such things as demand coverage, access to major roads, using existing parking sites, high performance transit priority as compared to bus system, proximity of each site to CBD, distance between sites, limited capacity for each site, locating sites beyond the congested area and etc.Past researches have identified criteria that are relevant to optimal park-and-ride location, but much of them do not address all of the above objectives simultaneously. And much of them have not assigned weight relationships to these criteria. For instance, various criterions may carry more or less significance than each other.In this paper a method for integrating these factors is presented simultaneously and a benefit based multi-objective optimization model including demand coverage, access to major roads, using existing parking sites, high performance transit priority as compared to bus system, proximity of each site to CBD in term of benefit(rials) is introduced. At the end of the research for showing the developed approach performance, the method is applied for sitting park-and-ride facilities in Mashhad, Khorasan Razavi, Iran.Steps in a case study are as: first we locate potential sites according to some of the objectives in the GIS environment and then with exported file as model Input data from GIS we will solving the introduced model using Genetic Algorithm and find the optimum location set among the potential sites.Using GIS is advantageous due to its vast information, storage, and graphical capabilities, and because of its major impact on the field of Location Science in terms of model application and model development is used in this research, and Genetic Algorithm was chosen for solving the model because of its ability to find the overall optimum solution in a spread and nonintegrated search space with lots of local optimum. The combination of these tools via an MATLAB interface allowed us to use geographical features of the urban network in our location process.Application results show the usefulness of the developed method in an urban region. One of the features of the approach developed here is that capacity of each site is limited at park-and-ride lots. While much of the existing methods and models do not address this objective. Besides in this research the objectives like coverage, access to major roads, proximity to CBD and construction of new sites are incorporated in a model simultaneously in term of cost, and the method can be used for sitting park-and-ride facilities in the context of an existing park and ride system. This could help transit agencies across the Iran (and abroad) aspiring to increase transit rider ship reach sound location decisions for their park-and-ride systems. As was shown in the paper, the model was applied for finding a compromise sitting configuration for park-and-ride facilities in Mashhad, Khorasan Razavi, Iran, from many available location set as a case study and the result of the model for the case that the target is sitting 5 locations in the city is showed.Solving the developed model with constant weighting method resulted in a unique solution, and using of weighting method for calibrating the model were useful for visualizing the effect of placing more importance on one objective relative to another. Moreover using weighting method is important for decision makers as they allow them to take the appropriate course of action(s).Issues for Further Study:During implementing this method as a case study, several issues inside and outside its original scope were identified. Many were addressed in this research; however, two issues should be further discussed as part of ongoing efforts and updates to this research.• finding or introducing a standardized approach to estimate benefits of objectives such as coverage, accessibility, proximity, transit modes priority and etc to get better model output.• Better understanding the park and ride facilities usage by Iranian private vehicles in order to modify the predicted capacity.