Navigation Kalman filter design for pipeline pigging
Navigation Kalman filter design for pipeline pigging
复制标题
DOI：
10.1017/s037346330500319x
复制
发表时间：
2005-05-01
影响因子：
2.4
通讯作者：
El-Sheimy, N
中科院分区：
文献类型：
--
作者：
Shin, EH;El-Sheimy, N
本文讨论了一种用于管道测量应用的自适应卡尔曼滤波的发展。建立了各种测量模型，并利用实际天然气管道数据集进行了数值测试。由于对于战术级动车组，仅靠里程计获得的速度测量不能产生良好的结果，因此增加了两个非完整约束来建立三向测量模型。平滑计算也被应用于使用过去、当前和未来测量中包含的所有信息来产生最佳轨迹。在极端情况下，轨迹误差在10-20米以内，在大多数情况下，轨迹误差在10英寸以内。轨道误差的均方根值预计在2.5米左右。
The paper discusses the development of an adaptive Kalman filter for pipeline surveying applications. Various measurement models are developed and numerically tested using a real natural gas pipeline dataset. Since, for tactical-grade IMUs, odometer-derived velocity measurements alone cannot yield good results, two non-holonomic constraints are augmented to make the three-direction measurement model. Smoothing computation is also applied to yield the best trajectory using all the information contained in the past, current and future measurements. The trajectory errors are within 10-20 m for extreme cases, and within 10 in for most cases. RMS of the trajectory errors is expected to be about 2.5 m.