Inertial Navigation System of Pipeline Inspection Gauge

Abstract

In this brief, a high-accuracy inertial navigation system (INS) for a pipeline inspection gauge (PIG) is proposed. Two mechanization approaches are investigated. First, a full INS dynamic model is used, and second, a 3-D reduced inertial sensor system (RISS) is utilized. The INS uses the full inertial measurement unit (IMU) data to calculate the navigation solution, whereas the RISS uses an encoder, one single-axis gyroscope, and two accelerometers. The RISS model is proposed in this brief for its better accuracy and less computational complexity. Due to the accumulated error in the INS or RISS solution, an extended Kalman filter is proposed to fuse the IMU data with PIG measurements. For the full INS model, these measurements are the encoder’s derived velocity constraint and the detected pipe length measurement. On the other hand, for the RISS model, it only refers to the detected pipe length measurement. An experimental setup, with a prototype of the in-pipe robot, is designed and built to test and validate our algorithms in a real pipe environment. Subsequently, the accuracy of the proposed algorithms is verified experimentally.