Adaptive tracking of maneuvering targets using two-stage Kalman filter

This paper presents a novel approach to the adaptive tracking of maneuvering targets using Kalman filter. The adaptation technique is achieved by constructing a pursuing system of two stages. The first stage comprises a vector Kalman filter of three states undergoes some biasing drift or mismodeling during a target evasive maneuver. While the second stage consists of a scalar Kalman filter basically developed to estimate the biasing term and then feeds back correction factors to the main processor to enhance its performance. This innovative adaptation technique will be called as bias removal (BR). Comparison and simulation exercises are presented against the interacting multiple model (IMM) adaptation technique showed that the target state estimates is improved by employing this BR adaptation policy.