APM(自动旅客捷运系统)轨道错台检测研究及应用

滕星权

Research and Application of Track Misalignment APM Detection Technology

TENG Xingquan
摘要:
[目的]为解决传统APM(自动旅客捷运系统)轨道错台人工检测方法的检测效率低、精度低等问题,需对APM轨道错台检测技术进行研究。[方法]利用线激光传感器获取轨道数据,基于多维图像处理技术和点云处理技术对错台进行识别和测量。对轨道轮廓数据进行图像化处理,通过图像处理技术识别轨道错台并定位,同时融合动态数列,判断在连续的时间段内图像识别算法识别的轨面状态是否一致,解决单个轨道断面误判的干扰,提高错台识别定位的准确性;错台精准识别和定位后,扫描获取轨道的点云数据,对数据进行滤波、平面拟合等处理,获取错台值的大小。[结果及结论]上海地铁浦江线的实际线路试验结果表明,采用多维图像处理技术和点云处理技术能够实现轨道错台的准确识别和定位,且识别的错台值的测量精度高、重复性好,满足轨道巡检要求。
Abstracts:
[Objective] In order to solve the problems of low efficiency and poor precision of traditional APM (automated people mover system), it is necessary to study the APM technology for track misalignment detection. [Method] The line laser sensor is used to obtain the track data, and track misalignment is identified and measured based on multi-dimensional image processing and point cloud processing technologies. Track contour data is firstly visualized, and image processing is then used to identify and locate track misalignment. Meanwhile, dynamic sequence is fused to determine whether the track status recognized by image recognition algorithm is consistent in a continuous period of time, thus solving the interference of single track section misjudgment and improving the accuracy of misalignment identification and positioning.After accurate identification and positioning of the misalignment, the track point cloud data obtained by scanning is processed by filtering and plane fitting to obtain the misalignment value. [Result & Conclusion] The test results from the actual line of Shanghai Metro Pujiang Line show that the multi-dimensional image processing technology and point cloud processing technology can realize accurate identification and positioning of track misalignment with high precision and good repeatability, meeting the requirements of track inspection.
论文检索