Accurate Pipeline Surveying Using Two-Filter Optimal Smoothing of Inertial Navigation Data Augmented with Velocity and Coordinate Updates

摘要

Pipelines are constructed to transport or dispose liquid and gases, commonly operated by oil, gas, sewerage and chemical industries. The pipeline system consists of three basic components namely gathering lines, trunk lines and distribution lines. Environmental, safety and economic concerns necessitate the constant monitoring of pipelines to avoid potentially hazardous failures. Currently, Pipeline Inspection Gauges (PIGs) can be sent through the pipelines to monitor the inside conditions. The Inertial Navigation System (INS) is employed to conduct the overall PIG navigation due to the unavailability of GPS signals inside the pipeline. Due to the time-dependent errors of INS only navigation, additional aiding sensors and/or auxiliary velocity and position updates need to be used to compensate for such errors. In addition, optimal smoothing methods are required since optimal smoothing is a post-mission estimator that provides the optimal estimates by using all available measurements. In this paper, two smoothers, namely Two-Filter Smoother (TFS) and the Rauch-Tung-Striebel Smoother (RTSS) will be implemented based on an Extended Kalman Filter (EKF). Using real pipeline inertial data collected with a tacticalgrade Inertial Measuring Unit (IMU), the effect of the additional velocity updates and available position updates will be shown. Moreover, the performance of both optimal smoothers will be evaluated. The combined implementation of the additional filter updates and smoothing demonstrated a remarkable improvement in the navigation accuracy compared to the original KF solution.