Analytical Bounds for an Interval Kalman Filter

This paper is concerned with analytical developments of results firstly introduced by the authors in [1]. These are devoted to optimization upper bounds interval covariance matrices appearing Interval Kalman Filter [2]. The proposed study mainly highlighted through two aspects. Firstly, further performed considering a class and minimizing traces these stages (in terms gain matrix then respect scalar parameter). Secondly, provides conditions under which optimal trace value controlled hence Algorithm [1], namely Optimal Upper Bound (OUBIKF), ensured perform stability (i.e. without width explosion resulting estimators). Also conditions, OUBIKF Algorithm, having similar structure Standard (SKF), get smaller bound correction step than one prediction step. Numerical simulations based on an automotive model illustrate developed results.