Window based Multiple Model Adaptive Estimation for Navigational Framework

Kalman filter based algorithms aim at providing accurate estimate of the state parameters which is indirectly governed by the accuracy of the sensor measurement and noise parameters fed to the system model. Multiple Model Adaptive Estimation (MMAE) is one of the adaptive techniques which tries to reduce the dependency of Kalman filter on the noise parameters fed to the system. The main goal of this work is to improve state estimation by incorporating window size as one of the unknown parameters in MMAE framework, referred to as Window based MMAE (WMMAE). The proposed scheme intertwines the concepts of Innovation Adaptive Estimation (IAE) and MMAE in one structure and the state estimation for each model is implemented by IAE. Simulation results prove the efficacy of WMMAE scheme as compared to MMAE and its other variants.