DefinePK

Abstract

Ammonia in palm oil fertilizer will affect to water pollution. Nowadays there are many types of fertilizer that have been and are being sold in Malaysia market. The low quality of fertilizer has the highest of ammonia quantity. By using one of significant instrument that was called as Electronic nose (E-Nose), the quantity of ammonia in palm oil fertilizer will be detected. Firstly e-nose chamber was developed by using Arduino, gas sensor array and potential meter. From e-nose with two types of ammonia in fertilizer samples, 2000 of data collected by microcontroller and spreadsheet software. By using mathematical method the data was normalized. Based on 2000 data, the graphical represented by MATLAB software. From the graph of two samples it is shown of two different pattern produced from the graph. Also by using MATLAB software and same data the boxplot will show the differences of data sample. After that, the normalize data was insert to Case-based Reasoning (CBR). CBR was employed by using excel software that was setup with mathematical methods. From CBR the 100% of accuracy, sensitivity and specificity have been achieved. As a conclusion, from the result of CBR, the project is successful to classify the different of palm oil fertilizer type.