作者: Renquan Lu , Shuwen Pan , Hongye Su , Jian Chu , Hong Wang
DOI:
关键词:
摘要: The problem of joint input and state estimation is addressed in this paper for linear discrete-time stochastic systems without direct feedthrough from unknown inputs to outputs. With the weighted least squares an extended vector including states, a recursive filter approach referred as Kalman with (KF-UI-WDF) derived. It shown that proposed KF-UI-WDF uniquely optimal sense both least-squares (LS) minimum-variances unbiased (MVU) over category MVU filters (e.g., [4], [5], [10]). global optimality also discussed. Due limited space, illustrative example omitted.