Coverage Method for the Location of a Light Rail Transit (LRT)

Due to the large number of urban trips and the importance of people's value of time, transportation planners are quite concerned about the design of efficient transportation systems. Therefore, the need for a new and efficient transit system for large cities is a must. Light Rail Transit (LRT) is one of transit alternatives for such cities. One of the important issues in the context of LRT development is network design. The core of the problem, which is designing a transit network capable of transporting a large number of people efficiently and cost effectively, is highly complex. In operations research, the problem belongs to the class of network design problems, but due to their very large sizes, integrality restrictions and non-linearity, they can not be solved by means of standard network design methods. There is not yet any analytical method available to solve this problem. Gendreau and his co-researchers have expressed the fact that design of a rapid transportation system is a procedure accompanied by complex trend of decision-making, multi-restraint and multi-objectives, uncertainty, non-measurable elements, heavy investment, and long period activity. They have then concluded that the standard traditional methods of network design are insufficient for such complex problems with big dimensions. Nonetheless, however, operation research methods might be of great importance and help at different stages of design, by developing and utilizing tools for production and evaluation of alternatives. Usually in practice, a two-stage procedure is used in locating the railway line, as following:1- Developing the efficient answer, considering the basic criteria, and2- Locating the best line amongst the existed alternatives by evaluating them and considering both measurable and non-measurable criteria.The studies conducted by Current (1985), Bruno(1998 and 2002), Dufourd (1996), Wirasighe (1997) and Laporte (2000); are all discussed and the relevant outcomes are fully considered in this study.In locating the transportation lines, those methods would be suitable to be employed, which possess the required capabilities of producing a number of suitable answers (and not only one answer) to be evaluated at the final step. In this regard those methods are preferred which produce a family of suitable answers based on the basic criteria. At the next stage, these answers would be evaluated based on other criteria, and finally the best answer would be chosen.The method employed in this study, is a two-stage procedure which is based on the explained view. For this purpose, at the first stage the covering-path method is used, and those corridors which potentially might contain LRT line, would be recognized. Then at the second stage, a generalized algorithm for the network case (combinations of single lines) is used. The objective at this stage is the recognition of the best combination. The basic idea in both stage, is the covering method for the defined benefits in the network.The criterions used in this study might be summarized as following: 1- Causing more demand for transit rail and promoting public satisfaction, 2- Minimum cost burden and maximum benefit-cost ratio,3- Maximum public access to the lines, and4- Transit line is expected to contain a systematic and coordinated relationship with points of potentially high trip production.Data needed for this study, are the results of the trips in 1385 (21 March 2006-20 March 2007), which are assigned on the existed transportation network in the city of Mashhad. All required information is provided by the Transportation and Traffic Organization of the city of Mashhad. Using econometric and trip demand models, the magnitudes of temporal trip demand are evaluated and then by utilizing EMME2 software, this demand is assigned to the network. EMME2 is the software for microscopic simulation of traffic volume. Thus, the volume of vehicles and trip time for passing each segment in each line is estimated. The mathematical model is totally developed by authors, and is solely used in this study. This model is also capable to be expanded for more objectives as following: Pj: the population assigned to node j of the network,Ej: the worker population assigned to node j of the network,(Vpub)ij: equivalent passenger car of transit volume at the arc (i,j) of the network,(Vpri)ij: equivalent passenger car of vehicle volume at the arc (i,j) of the network,(V(B-A)j): the volume of passengers of transit line at node j of the network,Cij: the cost of LRT construction related to link (I,j), Xij: a variable equal zero or one, indicating that the link is or is not on LRT line,d: the node at the end of the line,o: the node at the start of the line,A: the total of the links of the network,Aj: the total of the transit links of the network,Ap: the total of the transit links of the existed network,Ni: a sum of network nodes which are connected to node i with one link and node i is at the start of that link,Ni={k; (i, k) A}Mj: a sum of network nodes which are connected to node j with one link and node j is at the end of that link,Mj ={k; (k, j) A}The quantities of 1 to 6 in the objective function, show different benefits as well as the costs of LRT line construction. Restrictions 7 and 8 express the idea that the line starts at node zero and ends at node d. Relation 9 satisfies the required continuity of the links which constitute the line.The presented model in this study includes six objectives. This requires employing multi-criteria evaluation techniques to find the best alternative.The resulted algorithm was converted to visual basic and excel. The first one conducts step 1 and the second one. In this study the first and last nodes are purposely selected the same as the LRT line, which is under construction in the city of Mashhad. This enable us to compare the outcome with the results in previous study in Mashhad. In this study a practical and versatile heuristic method based on covering is applied to solve the problem of optimum network design of Light Rail Transit.