Run-Off Prediction Using Fuzzy Inference System

Abstract

An attempt of run-off prediction is carried out using an artficial intelligence system making use of the fuzzy theory in it's inference process. A new type of an artficial intelligence system which can treat “fuzzy sets” has been developed. The system consists of a fuzzy inference engine and a fuzzy knowledge base. Several rules for run-off prediction are placed in the fuzzy knowledge base. Run-off prediction simulation is tried out with real data from Nahari river area of Kohchi prefecture. Some promising results are obtained as an results of prediction simulation.