Recognition algorithms in E-nose: A Review

In recent years, the smart electronic nose (E-nose) has witnessed the rapid applications in diverse fields. Apart from sensor arrays, recognition algorithm plays a determinant role on the performance of E-nose. Focusing on the signal processing of E-nose, the response signal characteristic of a sensor is introduced first in this paper. Based on the differences between the processing of features, the algorithms are subsequently divided into traditional and artificial neural networks (ANN)-based, and their respective properties are specifically analyzed through the application in reality. The evaluation metrics for these algorithms are then summarized. Finally, the challenges and prospects of the algorithm are concluded. This paper aims to help researchers in diverse fields employ and explore the appropriate gas recognition algorithms for the emerging applications of E-nose.