Rancang Bangun Electronic Nose untuk Mendeteksi Tingkat Kebusukan Ikan Air Tawar

Abstract

metode integral sehingga didapat pola aroma sampel digital (digital fingerprint). Pola ini kemudian dianalisis dengan metode PCA (Principal Component Analysis) untuk menentukan pola aroma ikan air tawar. Hasil penelitian ini menunjukkan bahwa sistem electronic nose dapat mendeteksi aroma ikan air tawar dengan persentase variansi dua komponen utama sebesar 98,7% (bawal), 98,8% (lele), dan 99,5% (nila). Sensor yang memberikan respon tinggi pada setiap sampel adalah sensor TGS 2620 dan TGS 2600. Sensor TGS 822 memberikan respon tinggi pada ikan saat sudah tidak layak konsumsi. Kata kunci-Ikan, hidung elektronik, larik sensor, Principal Component Analysis (PCA). Abstract When fish die, fish freshness start to reduce gradually until cannot be eaten anymore. Properness of fish meat can be identified by odor that come out from fish itself. An instrument called electronic nose that can detect pattern of fish odor has been designed and implemented in this research. To be able to detect scent of freshwater fish, electronic nose will drain the air from sample chamber to sensor chamber using fan. When taking sample aroma, fan will drain air that contain sample scent from sample chamber to sensor chamber, and air from the outside flowed into sensor chamber when odor off. Scent stimulus captured by sensor array in form of signal response will be extracted with integral method so that the digital fingerprint from samples can be obtained. This pattern then analyzed by PCA (Principal Component Analysis) to determine patterns of freshwater fish odor. Result from this study indicated that electronic nose system can detect scent of freshwater fish with percentage variance of two major components are 98.7% (pomfret), 98.8% (catfish), and 99.5% (tilapia). Sensors that give high response in each samples is TGS 2620, and TGS 2600. TGS 822 give high response when fish is rotting.