Ethylene Gas Detection Using Neural Networks

摘要

In this paper, we consider the ethylene gas detection which is generated as one of by-products of the breathing in many plants. It may become harmful to the freshness of foods or vegetables since it accelerates a process to mature and cause self-acceleration of the deterioration. The density of ethylene gas in real world is extremely low. A gas chromatography and a mass spectrometry are used as measurement devices. Those machines are expensive and need much experience for operation. We use a neural network based on the features that are obtained from measured ethylene gases. Since it exists together with steam or alcohol in an environment, it is difficult to separate them each other. We make an odor measurement system of small size that is created by using the many sensors of crystal oscillators. Using the measurement system and neural networks, we made a detection device of the ethylene gas, selectively from water and alcohol.