Title: Fusion of dielectric technique and intelligence methods in order to predict acidity and peroxide of virgin olive oil

Authors: Mahdi Rashvand; Majid Javanmard; Abbas Akbarnia; Shahram Sarami

Addresses: Machine design and Mechatronics Department, Iranian Research Organization for Science and Technology, Institute of Mechanics, Tehran, Iran ' Iranian Research Organization for Science and Technology, Institute of Chemical, Tehran, Iran ' Machine Design and Mechatronics Department, Iranian Research Organization for Science and Technology, Institute of Mechanics, Tehran, Iran ' Biosystem Engineering Department, Iranian Research Organization for Science and Technology, Institute of Agriculture, Tehran, Iran

Abstract: Olive oil is one of the strategic and rich in minerals and nutrients among different oils. Due to the high price of olive oil, the quality of this product has a very important factor for consumers. Generally, the quality of olive oil is measured by two indexes of acidity and peroxide value. In this research, dielectric technique, artificial neural network (ANN) and support vector machine (SVM) methods were used to predict the acidity and peroxide value of olive oil. To analysis of output data in the range of frequency 1 KHz10 MHz, the artificial neural network with a topology of 1861-15-10 for acidity value and topology 1861-23-10 for peroxide value were predicted. Also, the best result of vector support was obtained by Gaussian algorithm with accuracy of 0.99. The results showed that the device and the evaluation methods were appropriate for prediction of acidity and peroxide value of olive oil.

Keywords: data mining; capacitive sensor; olive oil; quality.

DOI: 10.1504/IJPTI.2019.106196

International Journal of Postharvest Technology and Innovation, 2019 Vol.6 No.3, pp.192 - 202

Received: 26 Mar 2019
Accepted: 10 Oct 2019
Published online: 01 Apr 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article