Abstract
A novel recursive average triangulation modeling algorithm is proposed for the unknown but bounded noise systems. Since no priori knowledge can be used to denote the bounded noise term, the noise signal at each recursive step is warped in a strip with the hyperplanes obtained by the samples of input and output signals. The average triangulation idea is adopted and the minimum variance in the geometric space is defined as the criterion. In each recursive step, the number of simplexes changes with the variational triangulations that contain the true parameter values. The given simulations illustrate the feasibility and effectiveness of the proposed algorithm.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have