Sensor arrays and Self-Organizing Maps for Odour Analysis in Artificial Olfactory Systems
暂无分享,去创建一个
Our studies have regarded bio-inspired adaptive artificial olfactory systems composed of a sensor array for gas sensing and an artificial neural network, Self-Organizing Topology Preserving Map (SOM), introduced by T. Kohonen (Kohonen, 1989). In order to state the main working principles, Fig. I shows an overview of the digital version of a system for odour classification. The information flows from the left-hand side to the right-hand side: the gas mixtures in the environment determine the m sensor outputs which are sampled and converted into the digital stream z at each clock time. The module implementing the SOM network accepts a sequence of samples by a delay line, classifies the pattern according to its internal class models, and provides a class label as output (Davide et al.,1992 I,II,III).
[1] Riccardo Leardi,et al. A class‐modelling technique based on potential functions , 1991 .
[2] Olli Simula,et al. Process State Monitoring Using Self-Organizing Maps , 1992 .
[3] J. Gardner. Detection of vapours and odours from a multisensor array using pattern recognition Part 1. Principal component and cluster analysis , 1991 .
[4] Teuvo Kohonen,et al. Self-Organization and Associative Memory , 1988 .