列车定位轨道电子地图数据约简算法研究

摘要

高效、高精度的轨道电子地图是实现列车定位及列率运行控制的重要基础.为了提高轨道电子地图数据的存储及使用效率,在地图生成过程中需要在保证精度的前提下对测量数据进行约简.本文在实测数据的基础上分析轨道地图测量数据特性,采用小波变换及距离判别的方法对异常测量数据进行剔除,提出一种基于垂距判定和三次B样条曲线拟合的数据约简算法用于地图生成.该算法利用垂距判定方法从异常值剔除的原始数据中选取一定数目的测量点构成轨道特征点集,并对其进行三次B样条拟合,通过对拟合曲线进行误差判别修正和补充特征点,完成数据最终约简及轨道地图生成.青藏线实测数据验证结果表明,在合理选择垂距限差的情况下,该算法能够有效约简原始地图测量数据,并保证所生成地图的精度水平.%The electronic track map with high efficiency and precision is crucial for train locating and train operation control. In order to improve its storage and operating efficiency, it is necessary to reduce the huge measurement data with certain precision in track map generation. In this paper, the characteristics of the map measurements are analyzed with measured data, the wavelet transform and distance discrimination method are employed to reject abnormal measurement, and the vertical distance and cubic B-spline curve based algorithm is proposed for data reduction in map generation. In this approach, according to the vertical distance principle, some of the measurement points are selected as the feature points and used for B-spline fitting. With the error checking of the generated fitting curve, the feature point set will be supplemented and modified, then the data reduction and map generation will be realized completely. The measured data from the Qinghai-Tibet Railway demonstrate that with proper limited error of the vertical distance, the proposed algorithm can accomplish data reduction and map generation effectively and keep high accuracy.