Fuzzy Auto-regressive Model for Rainfall in Indonesia

Rain is one of the attenuation factors in very high frequency radio wave. Attenuation caused by rainfall gives a big impact in communication systems, especially wireless communication systems at millimeter-wave. This paper seeks to apply the fuzzy auto-regressive model (fuzzy AR) for rainfall in Indonesia. In this study, firstly the rainfall is measured, followed by getting the fuzzy variables for each event of rainfall measurements. Thirdly, the fuzzy variables based on the timing of the rain to get a fuzzy time series is set, and fourth step is the fuzzy time series modeling as a fuzzy AR(2) process. The fifth is estimating the matrix of model parameter, and the last step is simulating the generation of rainfall using fuzzy AR(2) model. The results obtained from the generation of simulated rainfall with Fuzzy AR(2) shows a good agreement with the actual measurement results.