管道惯性测绘原始数据迭代去噪算法  

The Arithmetic of Pipeline Geometry Mapping Original Data Iterative Denoising

作　　者：靳鹏[1] 杨理践[1] 高松巍[1] 

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870

出　　处：《无损检测》2016年第3期14-17,共4页Nondestructive Testing

基　　金："十二五"国家科技部支撑计划资助项目(2011BAK06B01-03);国家高技术研究发展计划"863"资助项目(2012AA040104)

摘　　要：管道惯性测绘内检测通常以搭载在清管器(PIG)内部的三维正交陀螺仪、加速度计,再结合里程计数据,解算出管道地理坐标轨迹。通常使用迭代卡尔曼算法进行原始数据去噪计算,但迭代的终止条件难以判断。利用管道发球筒及初始段管道平直,PIG初始姿态可控、可测的特点,结合静基座对准效果的评估,判断滤波效果,优化去噪过程,使初始对准及后续管道地理坐标的解算精度明显提高。Orthogonal gyroscopes,accelerometers,with odometer,are usually mounted inside the PIG for mapping by pipeline geometry PIG.The solution calculates the geographical coordinates of the pipeline track.An iterative Kalman algorithm was usually used to restrain the original data noise,but iteration termination conditions were difficult to judge.In this paper,because the initial segment of the pipeline and launch tube are straight,and PIG initial attitude is controllable and measurable,combined with the stationary base alignment effect evaluation,the filter performance can be determined and the de-noising process can be optimized,so that the initial alignment and subsequent pipeline geographical coordinates calculation accuracy has been improved significantly.

关 键 词：管道惯性测绘内检测 惯性测量单元 卡尔曼滤波 初始对准 

分 类 号：TG115.28[金属学及工艺—物理冶金]