Sistem Peramalan Cuaca dengan Fuzzy Mamdani (Studi Kasus: BMKG Lasiana)

Abstract

The weather is one part of human daily life. Many people who depend their lives on the weather to do every activity. Therefore, knowing the weather forecasting will give consideration to the community to be able to carry out various activities of human life such as in the field of aviation, shipping, agriculture, processed industries and others that depend on weather conditions. For this reason, the Indonesian BMKG has the duty to provide weather forecast information based on existing meteorological data using complex calculations. This study aims to build a system that will be an alternative for BMKG in forecasting weather using fuzzy based on four supporting criteria, namely air temperature, humidity, wind speed and air pressure. In doing weather forecasts using mamdani fuzzy there are several steps, namely determining the fuzzy set, the application of the implication function using the MIN function, the composition of the rules using the MAX function, and finally the Defuzzification process using the MOM method. This system will produce weather forecast results based on data on air temperature, air humidity, wind speed and air pressure that have been entered by the system user by showing the membership level of the predicted results. Based on testing that has been done, it is concluded that the system built using mamdani fuzzy can do a good weather forecast with a system accuracy rate of 61,062% using daily weather data as many as 1826 data in 2013-2017, with the lowest accuracy level found in 2015 with an accuracy rate of 54,247 % and highest accuracy in 2017 amounted to 65.207%.