Application of the Fuzzy Mamdani Method for Predicting Rainfall in the City of Yogyakarta

Abstract

Weather and climate play an important role in everyday life, influencing
various aspects of society and the environment. The city of Yogyakarta as a cultural
center cannot be separated from the impacts of climate change, including rainfall
patterns which can have significant effectiveness on daily activities, agriculture and
infrastructure. This research uses a fuzzy logic methodology approach with the
Mamdani method to predict rainfall. Fuzzy logic has proven to be effective in cold
situations and is able to overcome the challenges of missing or incomplete data. The
data used in this research was obtained from Bappeda DIY. The implementation for
predicting rainfall levels uses Matlab by utilizing the provided Fuzzy Logic Matlab
Toolbox. Based on research conducted using a fuzzy logic test to calculate all research
accuracy data, an average error value in RMSE was obtained of 39.88%. These results
indicate that there is significant error in these predictions, indicating that the fuzzy
logic control model is not completely accurate. Therefore, it is necessary to review the
data used in this prediction to ensure its accuracy.