Real Time Flow Prediction using Fuzzy Logic Models

Abstract
Models to predict flows in different parts of an urban catchment are developed using so-called Fuzzy Logic in combination with artificial neural networks. These models are developed to be used as part of a real time control system so that flows at various horizons can be computed and used as inputs to the controller. These models showed excellent performance both in their accuracy and execution time and proved to be very helpful in supplying the necessary forecasts for the real time controller. Another advantage of such models is that they can be developed in a relatively short time.
This paper is only available in PDF Format: View full text PDF