Design of Decentralized $H^{\infty}$ State Estimator in Indefinite Inner Product Spaces

부정 내적 공간에서의 준최적 분산 $H^{\infty}$ 상태 추정기 설계

In this paper, we propose a centralized $H^{\infty}$ state estimator for the multi state estimation problem using the result suboptimal $H^{\infty}$ filter is a special form of Ka filter whose state equations are defined in md inner product spaces. Con- ventional decentr filters are based on Kalman filter assumes precesses and measurements noises are w Gaussian noise. Therefore, Kalman based decent filter design hasn't robust performance in situation. Simulation results show that decent $H^{\infty}$ filter has robust perfotmance in worst case sensor fault situation.