Research On Vertical Optimization Of Urban Rail Line Using Genetic Algorithms

ABSTRACT:Accompanied by the acceleration of urbanization, construction of urban rail has entered a peak period. During urban rail design process, In the case of horizontal alignment has been determined, the vertical alignment directly affect the investment of construction and operation of a line. According to the energy consumption statistics of urban transportation operation, the locomotive traction fuel consumption accounts for around50to60percent of the total energy consumption. Therefore, conducting the study of profile optimization can bring practical social and economic benefits. Researchers at home and abroad have done a large number of studies about the optimization algorithm results, using the linear programming method, the gradient projection method, the dynamic programming method etc. These can better meet the needs of practical engineering. But in this process, researchers have less considered the operational energy. Therefore, through analysis and comparison of various kinds of methods, the optimization model based on the considering the engineering and operational costs has been established. And improved genetic algorithm has been designed to solve the model. Mainly works of this paper are as following:1. Established the model of traction calculation based on the multi-mass-point model and hybrid optimization strategy. Combined with the condition of railway location design parameters and subway train technical parameters, this project completed train operation simulation on the variant projects in railway location and drew the V-S curve. It provided necessary condition for solving the way to introduce the line train operation simulation conclusion into the profile optimization model.2. Then, ground line which was modified has been smoothed to determine the number of slope changing points automatically. The profile plan has been created by integrating uniform distribution method, fitting method and normal distribution method. Base on the hypothesis that the plan meet the demands of constraint condition and population diversity, initial population of the improved genetic algorithm have been formed. 3. Using the slope changing points, mileage and design elevation as variable of optimization design, satisfactory degree theory has been induced to establish profile optimization model, which can comprehensively consider both construction and operation. Base on analysis of connotation of profile design, multi-objective fitness functions have been established. Single slope fitness function has been proposed to draw into crossover operator, in order to moving direction of the slope changing points. Variation range of mutation operator and virtues or defect degree of the plan have been related to accelerate searching speed of the algorithm.4. Using the program to do optimization analysis for engineering application of a light rail AKO+000-AK18+125.405, the effectiveness and practicability of the program have been checked.