Weather forecasting in Kota Kinabalu by using Fuzzy Rules Based System

Tay, Fui Sing (2011) Weather forecasting in Kota Kinabalu by using Fuzzy Rules Based System. Universiti Malaysia Sabah. (Unpublished)


Download (8MB) | Preview


From ancient times until today, weather forecasting is an interesting and challenging task that has attracted uncountable researchers to involve in further study. Weather prediction uses various types of methods and weather parameters as inputs to produce a variety of weather events as outputs. Besides that, weather forecasting play an extremely important role to prevent the damage of disaster by announces the warning of weather phenomenon. Three main objectives of doing this dissertation are to perform weather forecasting by using Fuzzy Rules Based System, implement the system in Matlab to display the result in forecasting and to determine the accuracy of the system based on the level of error analysis in forecasting. In this study, two days scenario of four various types of weather parameters were observed to predict the next day event. This dissertation involved four type of weather parameter as inputs, included temperature, pressure, dew point and humidity and five chosen predicted weather events as outputs which are 'Sunny' event, 'Partly Cloudy' event, 'Rain' event, 'Thunderstorm' event and 'Rain-Thunderstorm' event. Consequently, 103 fuzzy rules were generated, then a Fuzzy Rules Based System was constructed and has successfully implemented in Matlab, to perform the daily weather forecasting along the year 2010 in Kota Kinabalu and for overall, the accuracy of prediction has achieved 68 percentages.

Item Type: Academic Exercise
Uncontrolled Keywords: weather forecasting, damage of disaster, Fuzzy Rules Based System, accuracy, weather parameters
Subjects: G Geography. Anthropology. Recreation > GE Environmental Sciences
Divisions: SCHOOL > School of Science and Technology
Depositing User: Unnamed user with email
Date Deposited: 02 Jun 2016 00:37
Last Modified: 30 Oct 2017 03:26

Actions (login required)

View Item View Item