Abstract

This paper presents an approach to time series extrapolation based on fuzzy control. The standard exponential averaging scheme is inflexible in that it gives a fixed weight to past history, thus ignoring transient phases in system dynamics. We present a modification to the scheme where the control parameter of the averaging scheme is dynamically adjusted by a simple fuzzy logic controller. The design of the controller is described, and the scheme is evaluated by simulation on test workloads and by application to the real-world problem of flow control in communication networks. We also study the sensitivity of our system to its descriptive parameters. 1 Statement of the Problem A frequently occurring problem in many areas of the physical sciences is that of extrapolating a time series into the future, given that the observed values can be corrupted by noise. Consider a scalar variable ` that assumes the sequence of values ` 1 ; ` 2 ; :::; ` k ; ` k+1 and can be represented as ` k+1 =...

Citations