Uncertainty Estimator

In this paper, we study estimation of system uncertainties in a dynamic system. We first modify the uncertainty estimator proposed in [1] so as to cover a larger class of system uncertainties; namely, those satisfying a "regular" condition. Then we propose an uncertainty-bound estimator, which estimates a magnitude upper bound of any "regular" uncertainty within a finite period of time. Combining this uncertainty-bound estimator with the original uncertainty estimator, we can achieve uncertainty estimaton with much less a priori information on the uncertainty compared with the original estimator in [1]. The results of this paper can be applied to problem such as controlling an uncertain system with minimal information on the system uncertainties.