Robotic simulation and GPS position estimation of a towed CRI equipment using Savitzky-Golay filter

Abstract

Because of the mobile nature of Capacitive Resistivity Imaging (CRI) towed arrays for subsurface surveys, the positions of its channels should be accurately recorded, as this is needed during the inversion process to create accurate resistivity maps. While CRI equipment that uses GPS to determine the positions is common and easy to implement, it is important to apply filtering methods to the obtained GPS coordinates to minimize the errors generated by the sensor. This study aimed to address the problem by applying the Savitzky-Golay filter to perform position estimation from noisy GPS measurements. The approach was tested in a virtual setup that uses CoppeliaSim for simulating the movement of the CRI towed array during a survey and a python script that records the GPS measurements from CoppeliaSim and applies the proposed filter to provide the position estimates. The script also generated sample resistivity maps given the noisy and estimated positions of the channels. Results showed that the Savitzky-Golay filter minimized each channel's positional errors during the 30 simulation runs to a mean error of 0.777m at most or a maximum of 61.62% decrease. Likewise, the importance of minimizing the position errors in generating the resistivity maps was proved as using the filter resulted in a more correct imaging of the subsurface for detecting the relevant regions and anomalies.