Bionic Optimization Design of Electronic Nose Chamber for Oil and Gas Detection

Abstract

In this paper, a miniaturized bionic electronic nose system is developed in order to solve the problems arising in oil and gas detection for large size and inflexible operation in downhole. The bionic electronic nose chamber is designed by mimicking human nasal turbinate structure, V-groove structure on shark skin surface and flow field distribution around skin surface. The sensitivity of the bionic electronic nose system is investigated through experimentation. Radial Basis Function (RBF) and Support Vector Machines (SVM) of 10-fold cross validation are used to compare the recognition performance of the bionic electronic nose system and common one. The results show that the sensitivity of the bionic electronic nose system with bionic composite chamber (chamber B) is significantly improved compared with that with common chamber (chamber A). The recognition rate of chamber B is 4.27% higher than that of chamber A for the RBF algorithm, while for the SVM algorithm, the recognition rate of chamber B is 5.69% higher than that of chamber A. The three-dimensional simulation model of the chamber is built and verified by Computational Fluid Dynamics (CFD) simulation analysis. The number of vortices in chamber B is fewer than that in chamber A. The airflow velocity near the sensors inside chamber B is slower than that inside chamber A. The vortex intensity near the sensors in chamber B is 2.27 times as much as that in chamber A, which facilitates gas molecules to fully contact with the sensor surface and increases the intensity of sensor signal, and the contact strength and time between odorant molecules and sensor surface. Based on the theoretical investigation and test validation, it is believed that the proposed bionic electronic nose system with bionic composite chamber has potential for oil and gas detection in downhole.