원문정보
초록
영어
Electronic nose (EN) is a equipment with ability of identification of simple or complex odors. Because of its low cost and accurate identification rate, the researches of it attract more attention and it develops quickly. Firstly, through Gas Sensor Array, we get large amount sample data and all data which is preprocessed. Secondly, modified Bp algorithm and RBF algorithm combining nearest neighbor - clustering algorithm and K-means clustering algorithm is proposed to realize the identification. Principal component analysis (PCA) method wipes off redundant sensor from the sensor array. Principle components and brief sensor signals are tested by above two algorithms. The test result indicated that the rapid and exact identification measure of ANN combining PCA is provided to the pattern identification with sensor information fusion.
목차
1. Introduction
2. Experiment System Model
3. PCA and Modified ANN Algorithm
3.1. PCA (Principal Component Analysis)
3.2. Modified BP and RBF Algorithm
3.3. Nearest K-means Clustering Algorithm
4. Experiment and Result Analysis
4.1. Data Preprocessing and Data Sample
4.2. PCA Research
4.3. Modified ANN Research
4.4. Result Reduce the Data Dimension of PCA
5. Conclusions
Acknowledgements
References