The Study Of Traffic Highway Network Planning Based On Genetic Algorithm

The development of the economy speeds the construction of highway Networks. The layout of the traffic networks resolves the path selection problem, and selects suitable time, direction and certain level to construct speedway. Its goals is to satisfy trends of traffic, have low affect on environment, and decrease constructing cost and other constraints.In general, a firm begins collecting all kind of materials about economy development and environment after accepting a assignment, and then studying a feasible path based on one of these materials individually. These suitable paths will be compared and outline a final solution.This traditional method cannot take into consideration global point of views. The solution has no attribution of science.Genetic Algorithms derives inspiration from the natural optimization process. The "survival of fittest" is applied to the population. They operate on encoded representations of the solution, equivalent to the genetic material of individuals. Moreover, they improve all potential solutions step by step through biological evolutionary process like crossover and mutation operation. The GA is very suitable for larger scale optimization problem. The latest population also gives a multiple solutions, and provides selection for strategy maker.This paper is based on Genetic Algorithms and gives a optimization solution for multiple constraints path selection problems, and provides foundation for strategy makers. This main study of paper centralizes on following aspect:1. based on Shortest Path Heuristic and Greedy algorithms, using path similarity principle and genetic operations, provides a heuristic algorithms in finding multiple feasible paths for multiple constraints (k-SPH).2. GA is used in traffic road networks plan firstly for finding multiple constrained paths.The experimental result for methods proposed in this paper is satisfied.The k-SPH algorithm uses mutation operation but not GA, and saves runtime.The paper is organized as follows: the first section describes the traditional path selection methods, shows all kind of constrained factors in planning traffic networks. The model and definition is described in the second section, and also provides general methods for path selection problems. In section 4,5,6, two methods for solving this kind problem is designed and algorithm is carried out. Conclusion is in the final section.