基于模糊聚类分析的高铁CPI稳定性判别

Stability Judgment of High-Speed Railway CPII Points Based On Fuzzy Clustering Analysis

运营期高铁精测网复测线上CPI点的稳定性分析是关乎最终CPII成果质量的关键一环,对不稳定线上CPI点的误判或漏判都会严重影响CPI平面成果,无法为轨道精调提供可靠基准,影响列车的行车安全。点位的稳定性分析是典型的模糊问题,稳定和不稳定的界限不明显,在无法保证线上CPI成果完全无误差的情况下,难以用限差精确划分稳定与不稳定。本文针对现有判断线上CPII点稳定性的方法不足之处,结合高铁线上CPI控制点复测更新判定指标具有的模糊性,引入模糊聚类分析,对点位变化量建立模糊相似矩阵,利用动态聚类图得到分类过程,计算变形水平值确定最优分类,并结合实测数据对聚类结果进行分析,证明此种方法在线上CPI点稳定性分析中的有效性。

The stability analysis of CPpoints on the retest line of high-speed rail precision measurement network during the operation period is a key part of the final quality of CPII results.The misjudgment or missing judgment of CPIlpoints on unstable lines will seriously affect the CPIlI plane results,and can not provide a reliable basis for track precision adjustment,affecting the train running safety.Stability analysis of point is a typical fuzzy problem,the boundary between stability and instability is not obvious,and it is difficult to accurately divide stability and instability with limit difference when CPllresults on line cannot be guaranteed to be completely error-free.In view of the shortcomings of existing methods to judge the stability of CPIpoints on the line,this paper combined with the fuzziness of the retest of CPIlcontrol points on high-speed railway lines,introduced fuzzy cluster analysis,established a fuzzy similarity matrix for the change of points,obtained the classification process by using dynamic cluster graphs,and calculated the deformation level value to determine the optimal classification.Combined with the measured data,the clustering results are analyzed to prove the effectiveness of this method in the stability analysis of CPllpoints on line.