摘要
轨道不平顺动态检测数据精准定位对识别轨道病害、挖掘不平顺演变规律从而提高现场养护维修效率至关重要。为修正检测数据里程与实际里程偏差并实现重复检测数据精准对齐,提出基于关键节点里程窗插值及相关优化整径(correlation optimized warping,COW)的里程校准三步法,利用最大相关系数累积值选取基准数据,并基于曲线、道岔台账里程窗插值方法修正基准数据绝对里程偏差;采用COW优化对齐算法对重复检测数据进行高精度对齐;基于历史病害信息对检测数据里程进行二次短窗绝对里程偏差修正。利用该方法对某长区段6次轨道几何检测数据里程进行修正,基于现场病害信息验证修正偏差,并运用相关指标评价对齐效果。研究结果表明,该方法可将检测数据绝对里程偏差降低至1.25 m以内,数据间对齐精度可达0.93以上,原始波形保持度达0.99以上,说明该方法可有效定位并对齐检测数据,从而提升线路轨道状态检测及病害识别定位的准确性。
Accurate positioning of dynamic track irregularity inspection data is essential for identifying track defects,uncovering irregularity evolution patterns and thus improving maintenance and repair efficiency.To correct the deviation between the mileage of inspection data and the actual mileage and to achieve accurate alignment of repeated inspection data,this paper proposed a three-step method for mileage calibration based on interpolation of key node mileage windows and correlation optimized warping(COW).First,with the use of the cumulative value of the maximum correlation coefficient to select the base data,the absolute mileage deviation of the base data was corrected based on the curve and turnout ledger mileage.Second,the COW optimized alignment algorithm was used to align the repeated inspection data with high accuracy.Third,a second short-window absolute mileage deviation correction was applied to the mileage of inspection data based on historical defect information.The method was used to process six times of track geometry inspection data of a long section,verify the mileage correction deviation based on the field defect information,and evaluate the alignment effect using relevant indexes.The results show that the method can reduce the absolute mileage deviation of inspection data to within 1.25 m,with the alignment accuracy between data of more than 0.93,and the original waveform retention of more than 0.99,indicating that the method can effectively locate and align the inspection data,thus improving the accuracy of line track condition inspection and defect identification.
作者
刁洪宝
张煜
张晓阳
杨飞
赵钢
DIAO Hongbao;ZHANG Yu;ZHANG Xiaoyang;YANG Fei;ZHAO Gang(Infrastructure Inspection Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;Department of Track,Communication&Signaling and Power Supply,China State Railway Group Co.,Ltd.,Beijing 100844,China;Infrastructure Inspection Center,China State Railway Group Co.,Ltd.,Beijing 100081,China)
出处
《铁道学报》
北大核心
2025年第2期111-121,共11页
Journal of the China Railway Society
基金
国家自然科学基金(52278465)
中国国家铁路集团有限公司科技研究开发计划(P2021T013)
中国铁道科学研究院集团有限公司科研项目(2021YJ022)。
关键词
轨道不平顺
里程偏差
窗插值修正
对齐算法
评价指标
track irregularity
mileage deviation
window interpolation correction
alignment algorithm
evaluation indicators
作者简介
第一作者:刁洪宝(2000-),男,山东平原人,研究实习员,硕士。E-mail:bbcess_1@163.com;通信作者:杨飞(1985-),男,山东枣庄人,副研究员,硕士。E-mail:13811807268@163.com。
