کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
1162873 | 1490913 | 2016 | 9 صفحه PDF | دانلود رایگان |
• Multifrequency large-amplitude pulse voltammetric electronic tongue was used.
• A visualized attributive analysis approach was created as an efficient tool for data processing.
• Rice taste flavor attribute was determined and predicted.
• The attribute characterization graph was represented for visualization of the interactive response.
This paper deals with a novel visualized attributive analysis approach for characterization and quantification of rice taste flavor attributes (softness, stickiness, sweetness and aroma) employing a multifrequency large-amplitude pulse voltammetric electronic tongue. Data preprocessing methods including Principal Component Analysis (PCA) and Fast Fourier Transform (FFT) were provided. An attribute characterization graph was represented for visualization of the interactive response in which each attribute responded by specific electrodes and frequencies. The model was trained using signal data from electronic tongue and attribute scores from artificial evaluation. The correlation coefficients for all attributes were over 0.9, resulting in good predictive ability of attributive analysis model preprocessed by FFT. This approach extracted more effective information about linear relationship between electronic tongue and taste flavor attribute. Results indicated that this approach can accurately quantify taste flavor attributes, and can be an efficient tool for data processing in a voltammetric electronic tongue system.
Schematic process for visualized attributive analysis approach using multifrequency large-amplitude pulse voltammetric electronic tongue for determination of rice taste flavor attribute. (a) sample; (b) sensors in electronic tongue; (c) excitation voltage program and response current signal from MLAPS; (d) similarity data matrix by data preprocessing and similarity extraction; (e) feature data matrix of attribute; (f) attribute characterization graph; (g) attribute scores predicted by the model.Figure optionsDownload as PowerPoint slide
Journal: Analytica Chimica Acta - Volume 919, 5 May 2016, Pages 11–19