基于可视化嗅觉技术的鲢鱼K值快速无损检测研究

Rapid Non-destructive Testing for K Values of Silver Carps Based on the Olfactory Visualization Technique

尝试采用可视化嗅觉技术对鲢鱼K值进行快速定量预测。利用可视化嗅觉技术对4℃恒温条件下不同冷藏天数的鲢鱼进行无损检测,获取可视化传感器阵列对样品顶空挥发性气体的响应信号;同时,利用高效液相色谱法检测鲢鱼体内三磷酸腺苷关联物的含量,算出K值;然后,采用偏最小二乘法(Partial least squares,PLS)和遗传算法偏最小二乘法(Genetic algorithm-partial least squares,GA-PLS)建立基于鲢鱼气味特征信息与K值的定量预测模型。结果显示,经遗传算法(GA)优化后原变量可从48个减少到18个,传感器可减少至11种;利用筛选出的变量建立的GA-PLS模型对鲢鱼K值的预测效果更好,预测均方根误差RMSEP=0.04,预测集相关系数Rte=0.93。研究结果表明,鲢鱼K值的实测值与预测值的相关性很高,可视化嗅觉技术能够用于定量预测鲢鱼K值。本研究为鱼类鲜度检测提供了一种准确、快捷、低成本的无损检测方法。

In this study, the olfactory visualization technique was used to predict the K values of silver carps quantitatively. The samples stored at 4 ℃ for several days were detected by the olfactory visualization technique non-destructively. The signals of the colorimetric sensor array respond to the volatile compounds were obtained. Meanwhile, the content of adenosine triphosphate related compounds in the samples was measured by HPLC and K values were calculated. Finally, the partial least squares （PLS） and genetic algorithm-partial least squares （GA-PLS） quantitative prediction models were established based on the odor characteristic information and K values. The results showed that the original variables reduced from 48 to 18 and the sensors reduced to 11 after the GA optimization. The performance of GA-PLS model was better than that of PLS. The root mean square error of prediction was 0.04 and the correlation coefficient of prediction was 0.93. The correlation between the measured and predicted K values of the silver carps was very high, therefore, the olfactory visualization technique could be used to predict the silver carp freshness quantitatively. This study provided an accurate, fast, low-cost and non-destructive testing method for fish freshness detection.