IDENTIFIKASI BUAH BERBASIS AROMA MENGGUNAKAN SENSOR GAS TERINTEGRASI KOLOM PARTISI

• Main Authors: , WAHYU SATRIA LITANANDA, , Ir. Sunarto Ciptohadijoyo, SU.
• Format: Theses and Dissertations NonPeerReviewed
• Published: [Yogyakarta] : Universitas Gadjah Mada 2014
• Subjects: ETD
• Online Access: https://repository.ugm.ac.id/129272/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=69662
• Institution: Universitas Gadjah Mada

In this research, an electronic nose was made based on combination of gas sensor and partition column. In the incorporation of these two methods, it would be proved whether using data from separated compound that can used as an input to the artificial neural network would be able to identify the type of the fruit that based on the scent, therefore it was done the basic research in identify the type of fruit based on the scent using gas sensor that integrated with partition column. The purpose of this research was to identify the type of fruit based from the aroma using an electronic nose based on the combination of gas sensor and partition column. Moreover, this research was also purpose to make software based on artificial neural network to identify the type of fruit that was built on a system. The material was fruits, that had a sting aroma, in this case, durian, jackfruit and pakel. For taking the data, it used the delphi serial comunication program, whereas for identifying the fruit was used backpropagation neural network. The length of data collection was set to be 5000 seconds, the heating temperature in the column was set to be 60°C and the gas of fruit which get into the column was set to be 90 seconds. The artificial neural network that was used in this research has structure [3 100 5 3]. The neural network was built with biner sigmoid activation function. The neural network used 3 input data from transformation result of difference peak value with baseline. The result showed that the artificial neural network was capable to identify the kind of fruits with accuracy 97.41 % from the variation of the tested data. Keywords: electronic nose, partition column, gas sensor, identification, fruit aromas, artificial neural network