Research Of Public Transportation Network Optimization Based On Heuristic Algorithm

In recent years, the economic construction of China’s major cities has made tremendous achievements. However, the lag of transportation infrastructure construction makes the sustainable development of cities encountered a bottleneck, so it is urgent to solve the problems of the urban traffic congestion. The optimization of public transportation network is the form of a "subtraction" to resolve this issue, compared to those "addition" operations such as extension of roads or increase of the traffic scale, it has small investment, more rapid and high efficiency, and a lower environmental cost to achieve the goal of the smooth flow of traffic, and more in line with the theme of "harmonious and sustainable development"This paper first analysis and summary the public transportation network optimization methods of domestic and international, and combine with the "respect history, respect for the status quo" practical requirement of this project, finally confirm the optimization idea of "combine with the reality, and gradually optimize". This method is based on the existing network structure, to improve the result by a continuous optimization, and ultimately achieve the desired degree of optimization. The goal of this paper is, under the premise of certain resources, by establishing a reasonable optimization model and designing an effective optimization algorithm, to optimize the existing network structure scientifically in order to get a better one, and finally achieve the purpose to improve the operational efficiency of the public transport system.Optimization model and algorithm are the core issues in the public transportation network optimization problems. On the basis of previous research results, and combined with the actual situation of the city, this paper set up a multi-objective decision-making model which contains a highest network coverage, a lowest route repeat coefficient and a least total travel time. This model takes route length and non-linear coefficient and the average station distance as the calculated parameters, and put forward an algorithm innovative which called "alternative route generation algorithm between two stations" as the neighborhood construction method, and combines with a passenger-sharing algorithm witch flow with the travel habits of the passengers to optimize the transportation network, and finally value those programs and choose the best one through the multi-objective evaluation system. On the logical control level of the algorithm, by analysis, summary and compare many of the heuristic algorithm, due to the higher implementation efficiency of the simulated annealing algorithm, both with its better robustness and the characteristic that the optimization results has nothing to do with the initial value, so we final choose the simulated annealing algorithm as the main control function.This paper takes the existing public transportation network structure of Yuxi as the starting point and the ultimate goal. We got the OD demand matrix of this city through on-the-spot investigation and data fitting, then put the data into the algorithm and got an optimized result, at the last we displayed and analysis the result. The data show that the optimized result in several important indexes are better than the original network structure, so as to prove the optimization model and algorithm we designed in this paper are feasible and effective.