基于遗传算法的铁路纵断面优化系统的研究

Study on Railway Profile Optimizing System based on genetic algorithms

提出一种基于全局最优的改进式遗传算法。采用改进的启发式交叉算子和多种算子混合操作,加快算法向最优解收敛的速度。给出相应的纵断面优化模型,并改进初始种群生成的方法,将一个二维优化问题转化为两个一维优化问题进行求解,简化了算法的执行过程。引入惩罚函数的概念处理方案的约束问题,设计算法实现的流程图,并开发相应的优化系统。利用实例证明遗传算法在纵断面优化设计中的可行性及全局寻优的性能。

It was put forward the modified genetic algorithm. A mixed operation between improved heuristic crossover operator and multi-operator was adopted to accelerate the convergence of the algorithm to the optimal solution. The corresponding model of vertical section optimization was put forward. The method to improve the method of generating initial population, by transforming a two-dimensional optimization problem into two one-dimensional optimization problems, was also improved. The constraint problem of program was dealt with by utilizing the concept of penalty function in this paper. A flow chart of the algorithm was designed. The Corresponding Optimization System was developed as well. The example adopted in this paper approved the feasibility and the performance of global optimization of genetic algorithms to optimize the design of vertical section.