Taste Recognition of LAPS Electronic Tongue Neural Network

JIANG Xing-guo,CHU Fu-gang,CHEN Zhen-cheng,LIANG Jin-tao
DOI: https://doi.org/10.3969/j.issn.1000-3428.2012.13.007
2012-01-01
Abstract:According to the Light Addressable Potentiometric Sensor(LAPS) principle,a taste recognition system is proposed based on algorithms of Principal Component Analysis(PCA) and Neural Network(NN).Using PCA to extract principal components of the original data,and the principal components are used as the training samples,to find the intrinsic links between the taste data and the taste type through learning and training process.The Back Propagation(BP) network trained is used to recognize the taste.Tastes of acid,sweet,bitter,salty and umami under density of 20 ppm,100 ppm,300 ppm and 500 ppm are recognized respectively.The rate of correctness is up to 96.6%,which shows that the taste recognition system can be used to identify the taste of liquid when the density is changed.
What problem does this paper attempt to address?