Adaptive Modified Wave Estimator

Kalman filter is frequently used for integration of the navigation systems. Process noise variance, employed in the calculation of the Kalman filter´s state prediction covariance, determines error estimation capability of the filter for navigation system. Due to the difficulties in exact modelling, i.e., determining the exact value of the process noise, Kalman filter´s performance could become limited. Recently, modified wave estimator (MWE) has been suggested for the state estimation of especially weakly observed states with high accuracy. Unfortunately, due to cycle time calculations, computational burden of the MWE is very high. In this paper, adaptive modified wave estimator is suggested in order to overcome the computation issue. Estimation performance and computational burden of, Kalman filter, MWE and AMWE are discussed for a selected navigation application