Hybrid Algorithm Based On Genetic Algorithm And Dynamic Programming Railway Profile Optimization Design

Railway alignment design is the most important part of railway design. On the basis of compliance, Control of project investment and the size of the bill of quantities is the design main goal of track design, which is the important purpose of route optimization. Based on the optimization theory, in the case of horizontal alignment have been confirmed, Looking for profile scheme based on minimum quantities.At first, using mileage and elevation of grade change point as its design variables, using fill and cut area of the centerline of the profile as objective function, and considering various constraint effects, this paper established a mathematical model of railway profile. Smooth the original ground line is the first step to generate the initial profile. The point of inflection of different smooth curve of curve radius is recognized as inflection point, the maximum rough degree of points of inflection point on both sides of the smooth curve as the fitting region. Using the least squares fitting slope line in turn, intersection point of grade line is the initial grade change point, Through constraint processing can get the initial profile. For the genetic algorithm optimization model,In this paper, the program automatically generate profile and existing profile as the initial optimization solution, and relevant profile optimization models about genetic algorithm、dynamic programming and combine two kinds of algorithms. For the genetic algorithm optimization model, in order to prevent premature convergence problem of iteration’s results, more than one operator and add new profile program after the completion of each iteration are adopted to solve this problem. Dynamic programming optimization profile can ensure the optimal results of each phase and the optimum result of global at the same time, in order to keep balance of fill and cut, With the previous amount of fill and cut and the current amount of fill and cut as the phase’s optimal function.Hybrid algorithm is less applied in railway line optimization. Combination of two algorithms to optimize the profile not only make full use of the advantages of single algorithmand can make up for the shortage of between each other. By using global search of genetic algorithm and dynamic programming’s characteristics of small scale search and local optimization to find the optimal solution within the allowable scope.This paper uses matlab software write relevant processes. There are mainly four parts(profile program generator automatically, genetic algorithm optimization procedure,dynamic programming optimization program and optimization program of combining algorithm). Through an engineering example compare the amount of fill and cut of before and after optimization profile. The algorithm procedure is effective and feasible by verifying, and there are practical significance.