基于电子鼻和非线性数据特征分析的鸡肉鲜度检测方法∗

摘要

Based on electronic nose and non￣linear data feature analysis,chicken meat freshness detecting method has been investigated. Chicken chest meat is selected as samples,and stored at 4 ℃ for continuous 6 days. Each day,TVB￣N index and electronic nose responses are measured. Results demonstrate that samples stored over 3 days are not fresh according to China standard protocols. Principal component analysis( PCA) and non￣linear stochastic resonance( SR) are used to process electronic nose measurement data. Results indicate that PCA method only quali￣tatively discriminates samples stored under different storage time. SR outputs cross correlation coefficient( CCC) dis￣criminates all meat samples. Chicken meat freshness detecting model is built based on CCC eigen peak values. The proposed method presents some advantages including rapid analysis,good accuracy,low cost,etc. It is promising in meat quality analysis.%探讨了一种基于电子鼻和非线性数据特征分析的鸡肉鲜度检测方法。以鸡胸肉作为样品，样品在4℃温度下连续存放6天，每天检验样品的挥发性盐基氮( TVB￣N)指标，并且实验检测鸡肉样品的电子鼻响应数据。实验结果表明储存了3 d的鸡肉样品TVB￣N指标超出了国家标准限定值而无法食用。采用主成分分析( PCA)方法和非线性随机共振方法分析电子鼻检测数据，结果表明主成分分析方法仅能定性的区分不同储存时间的鸡肉样品，随机共振系统输出相关系数曲线可以较好的区分不同鸡肉样品，并且依托互相关系数特征峰值构建了鸡肉鲜度分析模型。该方法具有检测快速、准确性好、成本低等优势。