Distributed identification of fuzzy confidence intervals for traffic measurements

A distributed fuzzy confidence interval identification method for traffic measurements is considered. Using historical data of the density measured on different sections of the freeway, the idea is to find fuzzy confidence intervals that define the bands that contain almost all the density measurements. The purpose of the proposed approach is twofold. First, to obtain a band as narrow as possible for each of the sections of the freeway. And second, to have a high percentage of the data contained in the bands. The method we propose in this paper is completely distributed, and can be used not only to describe any uncertain nonlinear distributed parameter system but also as a key element in a robust controller.