Chinese bayberry (Myrica rubra Sieb. et Zucc.) quality determination based on an electronic nose and non-linear dynamic model

Abstract

In this paper, a Chinese bayberry (Myrica rubra Sieb. et Zucc.) quality determination method using an electronic nose (e-nose) and non-linear stochastic resonance (SR) technique has been studied. E-nose responses to bayberry samples stored at 4 °C for 7 days are measured. In order to characterize the sample quality physical–chemical indexes, such as human sensory evaluation (HSE), texture, color, pH, total soluble solids (TSS), and reducing sugar content (RSC), are examined. The e-nose measurement data is processed by principal component analysis (PCA), SR and double-layered cascaded series stochastic resonance (DCSSR) methods. PCA can not totally discriminate all bayberry samples. Bayberry SNR maximum (SNR-Max) values calculated by SR and DCSSR increase with an increase of storage time. SNR-Max values successfully discriminate all bayberry samples. Measurements based on multiple variable regression (MVR) between physical–chemical indexes (firmness, pH, color, TSS, and RSC) and SR/DCSSR SNR-Max values have been conducted. Results indicate that SR is more suitable for Chinese bayberry quality determination compared to DCSSR. The bayberry quality predicting model is developed based on linear fitting regression of SR eigen values. The validation experiment results demonstrate that the developed model predicts bayberry quality with an accuracy of 95%. The proposed method has many advantages including easy operation, fast responses, high accuracy, good repeatability, and low cost.