Fuzzy Expert Forecasting System for Coconut Scale Insect Infestation
Juliet O. Niega, Bryan L. De Guzman, Joseph P. Alcoran, Frederick R. Dalena, Marlon Aves Diloy, Leandro R. De Luna · 2022 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) · 2022
The purpose of this study is to develop a coconut scale insect infestation forecasting system for Philippine Coconut Authority in Southern Luzon, Philippines. The system was developed using Sugeno-style fuzzy inference in MatLab and realized in MS Excel VBA Macro. The study examined the effect of weather parameters such as temperature, relative humidity, and wind speed on coconut scale insect infestation and predicted the potential future degree of infestation. The Fuzzy Inference System technique applied in this study includes the use of a triangular membership function, the formulation of 27 If-Then rules, the use of AND logic operator, and the weighted average defuzzification. The designed system was tested by comparing recorded data from the 2014 CSI infestation in five provinces of Southern Luzon, Philippines, to the data generated by the developed software. The actual infestation and the data generated by the developed GUI differed by 1 to 5%. Experts believe that this is due to the government's intervention to address the CSI problem. Furthermore, temperature has a 45% impact on the infestation, followed by wind speed (30%) and relative humidity (25%). The overall system evaluation result is 4.1, which is interpreted verbally as very good. As a result, it was concluded that the applied fuzzy logic concept, which is developed into VBA Macro, is a useful tool for forecasting the severity of CSI infestations.