This paper proposes a compensated robust Kalman filter to reduce estimation error of a practical system. The practical system is assumed to have system uncertainty and affected by external disturbance and noise. Proposed Kalman filter has a compensation input, which reduces discrepancy between error dynamics of the reference system model and that of the practical system. The compensation term is calculated with LMI (linear matrix inequality) by comparing estimation errors of two system. Simulation results show that estimation error of practical system reduced by applying proposed robust Kalman filter.
