摘要
归纳和总结了铁路线路智能优化与三维空间信息之间的内在联系,基于OSG技术对空间信息数据进行高效组织,加工处理与信息输出为一体,建立了三维空间信息模型,为铁路三维空间线路智能搜索提供可视化地理信息环境。以平面交点坐标、交点半径、纵面变坡点里程、变坡点高程为设计变量,充分考虑了空间线路平面约束、纵断面约束、平纵组合约束和环境影响约束条件,深入分析铁路三维空间线路优化费用目标函数,建立了铁路三维空间线路综合优化数学模型。采用浮点编码方式,以交点偏移距、交点曲线半径、链式变坡点高程为基因序列,针对多约束条件构成的优化空间进行深入的研究,生成线路方案群;基于多目标排序矩阵方式对每代中线路方案进行适用度计算,设计了选择、交叉和变异三类遗传算子,逐代遗传进化,实现了线路方案向最优线路方案群自动搜索,完成了铁路三维空间线路智能寻优过程。以本文提出的理论与方法为基础,基于vs.net、OSG、数据库等技术实现了铁路三维空间智能选线系统的开发,结合实际工程对本文的理论模型与算法进行了验证和评价。
The internal relationship between the railway line intelligent optimization and three-dimensional geo-spatial data is szcmmarized. Based on OSG, it firstly establishes the three-dimensional spatial information model by organizing, processing and outputting the spatial data as a whole, which may provide visualization of geographi-cal information environment for searching the railway line schemes in three-dimensional space. Then, it analyzes the cost objective function of the railway lines and builds up the comprehensive optimization model for railway line intelligent optimization in three-dimensional space. Besides, it calculates the applicability of line schemes based on multi-objective sorting matrix and designs the genetic operators of selection, crossover and mutation. The line schemes to be near the group of optimal route scheme is realized by genetic evolution. And railway line intelligent optimization in three-dimensional space is completed by searching performance, forming the group of optimal and valuable line schemes. Based on the theory and method which are proposed, the intelligent line selection system in three-dimensional space is realized based on the technology of vs.net, OSG, database technology, etc. Finally, the verification and evaluation of theoretical model and algorithm has been carried out by using the practical engineer-ing projects.
出处
《华东交通大学学报》
2014年第2期48-61,共14页
Journal of East China Jiaotong University
关键词
铁路选线
智能优化
改进遗传算法
空间信息模型
多目标
数学优化模型
railway location
intelligent optimization
improved genetic algorithm
spatial information model
multi-objective
mathematical optimization model