A dead reckoning localization method for in-pipe detector of water supply pipeline: an application to leak localization

Abstract

Urban water supply pipeline system integrity is important for the urban life. The aim of the study reported in this paper is to locate the water pipeline leaks by using an in-pipe detector. In this study, a mathematical model is extracted from an actual inspection system. By using the homogeneous transformation theory, transformation matrix which is from carrier to a reference coordinate system is deduced, and then the global transformation matrix is obtained to describe the detector’s posture. Through measuring the distance increment of each sample time step in carrier coordinate system, the cumulative distance result is calculated. After combining the data of the inertial measurement unit (IMU) and odometer, the leak can be located. To improve the accuracy of leak localization, the magnetic markers are implemented about one in each 1 km distance, which provide reference points to be used to compensate accumulative error during the localization process. Furthermore, a dead reckoning localization method combining data of a micro electro-mechanical IMU, three odometers, and magnetic markers is proposed. To verify above localization algorithm, a simulation case study is conducted with the artificial error generated by the white noise. The simulation results show that the dead reckoning algorithm can effectively provide leak locations with a reasonable uncertainty. Based on this, an experimental platform was built in this study. The experimental results show that the relative error of leak locating achieves a reasonably good performance