Design of an EKF-CI based sensor fusion for robust heading estimation of marine vehicle

An efficient approach for deriving accurate pose and heading values through multi-sensor fusion of data from several inexpensive sensors (such as multiple GPS (Global Positioning Systems), EC (electronic compass), rate gyro) is presented. The proposed multisensor fusion approach is composed of several sub-methods namely initial heading calculation, classification and weighing (CnW), extended Kalman filter (EKF) and then covariance intersection (CI) algorithms. The consecutive implementation of the sub-methods gives an accurate heading value with lesser RMSE (root mean square error) compared to the original GPS COG (course over ground) and EC. Several experimental tests were done to confirm the good performance of the proposed process.