Compensation of Magnetic Interference for Precise Directional Control of Drones

Abstract

Magnetometers are essential for directional control of drones. This paper proposes an eff icient algorithm for compensation of magnetic interference induced by the current owing in brushless DC (BLDC) motors of drones in order to achieve their precise directional control. To begin with, a state space model is constructed to involve a gyroscope as well as a magnetometer with its external interference. Based on the constructed state space model, the Extended Kalman Filter (EKF) is designed to estimate the direction of a drone by using measurements from two sensors, or the gyroscope and the magnetometer. By adjusting the measurement noise variance of the EKF according to thrust of BLDC motor, magnetic interference can be effectively rejected to increase the accuracy of direction estimation. In addition to direction estimation through a sensor fusion algorithm, online calibration is done periodically to update the elliptical locus of its measurements and hence reduce interference. The practical effectiveness of the compensation scheme for attenuating magnetic interference is illustrated through real experiments with a drone.