Carrier Magnetic Interference Compensation Method Considering Dynamic Magnetic Bias Caused by ROV Thrusters

Abstract Remote Operated Vehicle (ROV) equipped with scalar magnetometers is a typical method for detecting magnetic targets. The entire system framework is usually designed to be relatively compact to maintain a stable attitude, however, these compact frameworks require additional consideration of the dynamic magnetic interference generated by the ROV's power system, which reduces the measurement accuracy. Dynamic magnetic bias resulting from fluctuating currents can be mitigated by either positioning the magnetometer away from sources of bias or employing expensive design solutions and practical experience. Neither option is feasible within a compact and cost-effective system framework. Consequently, this paper proposes a novel compensation method for carrier magnetic interference that accounts for dynamic magnetic bias. A linear model is established by estimating the coefficients that map throttle command to time-varying magnetic interference. Together with permanent and induced magnetic interference, a comprehensive compensation model for carrier magnetic interference is formed. Building on this foundation, a new state estimator based on the interpretable framework of the Kalman Filter (KF) is utilized to estimate the compensation parameters. This approach implicitly learns the Kalman gain from data, thereby circumventing the KF's reliance on underlying noise statistical knowledge. Field experiment shows that this method significantly improves the compensation effect of magnetic interference.