Sensor Error Analysis of Superconductive Angular Gradiometer

Abstract

A superconductive angular gradiometer is a pair of superconductive angular accelerometers that pivot about a common axis. The output of the sensor corresponds to gradients. The gradients are effective not only in explorations of ore bodies and oil fields, but also in a complete modeling of the gravity field of the Earth. The superconductive angular gradiometer typically carries by a mobile platform to collect the measurements. Therefore, the instrument senses not only gravity gradients, but also various effects of an accelerated coordinate frame (introducing errors). The errors in the gradiometer channel are technically removed through some electrical, mechanical, and empirical approaches. A major effort in this study is given to an efficient and a cost effective post processing approach to deal with only the effect of angular velocities squared.

A superconductive angular accelerometer is an important sensor that often supplements a mobile superconductive angular gradiometer during data acquisition operations. While the angular accelerations are measured, the angular velocities squared are computed by integration. However, the angular accelerations are noisy because of translational accelerations of the platform and temperature fluctuations of the environment during the operation. Wavelet de-noising and de-trending techniques have been implemented in order to mitigate these errors. The results indicate that more than 66% of the noise level was efficiently reduced in comparison to the empirical approach by the company for the static x angular accelerations. This improvement was achieved without knowledge of the temperature and other error effects.

Furthermore, the effect of angular velocities squared, which is about -20 Eötvös, is removed from the gradiometer output. Then, the performance of the gradiometer is evaluated through a simulation study by error propagation in a single-input-single-output system. The standard deviations of the static x angular acceleration, y angular acceleration, and differential mode after the wavelet analysis are 3.08e-05 rad/s2, 3.55e-05 rad/s2, and 3.21e-08 rad/s2, respectively. So, the the final accuracy of the true gradient in this research work is 0.06 Eötvös, which makes the system suitable for variety exploration applications that require the gradients to be known with an accuracy of 1 Eötvös or better.