This paper addresses the state-estimation problem for nonlinear systems for the case in which prior knowledge is available in the form of constraints on the state vector. A review of methods investigated in the literature is presented together with new results. The main contribution of this paper is to categorize the investigated constrained state estimators in five groups according to the way in which the constraint information is enforced, namely, measurement augmentation, estimate projection, sigma point projection, quadratic programming, and truncation. Two illustrative examples are discussed.
Kalman filter; state estimation; constraints; unscented Kalman filter
