Derivation of Modified Input Estimation Filter Using Bayesian Framework

Abstract

Abstract A new approach to derive the modified input estimation is proposed using Bayesian framework. Tracking manoeuvring targets is a tough problem and many algorithms are proposed to deal with it. Among these algorithms, modified input estimation (MIE) technique has been proved to be an effective method. MIE technique treats acceleration as additive input term in the state vector and estimates the original state and acceleration vectors simultaneously with a standard Kalman filter. Analogous to standard Kalman filter, however, the MIE filter equations which have a modification on Kalman gain is given without a cohesive derivation. Using a Bayesian framework, we will present a conceptually cohesive roadmap that starts at first principles and leads directly to derivation of MIE filter equations.