Improvement of E-Nose Sensor Signal Using MVA, FFT, DWT Methods on Pineapple Fruit Maturity

In this study, we built an Electronic Nose to detect pineapple ripeness. We designed a prototype E-Nose using a wooden crate. Inside the box, we built the E-Nose using 9 MQ sensor circuits connected to the Arduino Microcontroller. In the crate, we put an object, namely a pineapple with three levels of ripeness (ripe, half ripe, and unripe), each weighing 1kg placed in three different positions (5cm, 20cm, and 35cm) from the position of the sensor array (E-nose). We converted the signal output results into ppm units. We compared the value of each fruit signal based on the level of ripeness and distance using E-Nose. We used several signal processing methods (wavelets) for signal processing. Then, we improved the signal generated using wavelet. The wavelet methods used are Moving Average, Fast Fourier Transform (FFT), and Discrete Wavelet Transform (DWT). The contribution of this research is to see the influence of the signal valuegenerated from the sensor (MQ Sensor) to the position (distance) of the object (fruit) and its influence on the level of maturity (fruit). Then signalgenerated from the sensor we make improvements with the mva, fft, and dwt methods.