Improving multiple model adaptive estimation by filter stripping

State estimation for a given mathematical model of a system is one of the core frameworks which is used to predict state at every instant. These schemes act as noise filters and are able to substantially reduce their effect over measurement. Kalman filter is one of the most widely schemes in this domain whose accuracy is indirectly governed by the accuracy of process and measurement noise parameters fed to the system. Innovation Adaptive Estimation (IAE) and Multiple Model Adaptive Estimation (MMAE) are two main forms of adaptive scheme which is able to reduce the filter´s dependency on these parameters. In this work, MMAE is studied and improved by incorporating a novel concept of stripping down several Kalman Filter (KF) blocks to a limited number. The proposed methodology has been compared with other improvements carried out in MMAE framework and is presented for analysis and detailed discussion.