Optimization of Sugeno Fuzzy Inference System’s Rules for Precipitation Predictions

Abstract

Precipitation information and forecasting is very important for many fields such as airport management, shipping, soil irrigation, flood control, water reservoirs and distribution. However, obtaining an accurate forecasting result is not an easy task. The Sugeno Fuzzy Inference System can be used as the prediction method. To obtain more accurate results, rules in the method are optimized using Evolution Strategies Algorithm. This approach is compared with Generalized Space Time Auto Regressive (GSTAR) Model. Several phases of numerical experiments show that the optimized Sugeno Fuzzy Inference System obtains better results comparable to those achieved by Generalized Space Time Auto Regressive.

GS Cited Analysis