Intelligent classification of waxy crude oil odor-profile at different temperature

Abstract

Crude oil is one of the basic needs required for humans to ease their life. The quality of crude oil with the lowest wax content is very important, in order to sustain the transportation and production of crude oil from offshore to onshore. Based on literature from previous studies, the appearance of wax depends on the temperature which is called Wax Appearance Temperature (WAT). Hence, there is a need to propose a new method to classify the waxy crude oil at a different temperature. The main purpose of this paper is to classify Malaysian waxy crude oil odor profile at different temperatures using intelligent classification technique. There are 28,000 data measurement of the waxy crude oil that was taken using an electronic nose (E-nose). The data readings have been normalized and analyzed using a statistical method. Then, the odor profiles were classified using K-Nearest Neighbour. The classification performance shows that the technique was able to classify the Malaysian waxy crude oil odor profile at different temperatures with 100% accuracy.