期刊
TALANTA
卷 162, 期 -, 页码 218-224出版社
ELSEVIER SCIENCE BV
DOI: 10.1016/j.talanta.2016.09.055
关键词
Electronic tongue; Voltammetric sensors; Partial least squares regression; Wine; Sensory panel; Ageing
资金
- Spanish Ministry of Economy and Innovation, MINECO (Madrid) [CTQ2013-41577-P]
- Premier's Research and Industry Fund through Catalyst Research Grants [CRG69]
- Universtitat Autonoma de Barcelona for the PIF fellowship
- ICREA Academia
This work reports the application of an electronic tongue as a tool towards the analysis of wine in tasks such as its discrimination based on the maturing in barrels or the prediction of the global scores assigned by a sensory panel. To this aim, red wine samples were first analysed with the voltammetric sensor array, without performing any sample pretreatment. Afterwards, obtained responses were preprocessed employing fast Fourier transform (FFT) for the compression and reduction of signal complexity, and obtained coefficients were then used as inputs to build the qualitative and quantitative models employing either linear discriminant analysis (LDA) or partial least squares regression (PLS), respectively. Satisfactory results were obtained overall, with a classification rate of 100% in the discrimination of the type of barrel used during wine maturing, a normalized NRMSE of 0.077 in the estimation of ageing time (months) or 0.11 in the prediction of the scores (0-10) from a trained sensory panel (all for the external test subset).
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据