便携式水产品多品质参数拉曼检测装置设计与试验

Development of Portable Device for Detection on Multi-quality Parameters of Aquatic Products by Raman Spectroscopy

水产品在捕捞后极易发生腐败变质,传统的品质检测系统体积大、移动不便、操作复杂,不利于实际应用推广。为了满足水产品品质快速、实时、无损的检测需求,基于拉曼光谱技术,研发了便携式水产品多品质参数检测装置。该装置的硬件系统主要包括光源模块、光谱采集模块、系统控制处理模块、电源模块、通信模块。采用标准正态变量变换、Whittaker平滑算法、自适应迭代重加权惩罚最小二乘算法对该硬件系统采集的拉曼光谱数据进行平滑和背景扣除处理,结合国标方法采集到的鲳鱼颜色b*、硫代巴比妥酸标准理化值(TBA值)、挥发性盐基氮(Total volatile basic nitrogen,TVB-N)含量,建立了鱼肉品质多指标偏最小二乘定量预测模型。鲳鱼颜色b*、TBA值、TVB-N含量的验证集相关系数分别为0.907、0.897、0.915,验证集均方根误差分别为1.315、0.104 mg/(100 g)、2.875 mg/(100 g)。基于MFC基础类库完成实时分析控制软件设计,将预测模型植入软件内,交叉编译下载到检测装置中,实现水产品多品质参数指标一键检测。最后对装置的稳定性进行了测试,18条鲳鱼样品的颜色b*、TBA值、TVB-N含量的装置预测值与标准理化值的相关系数分别为0.927、0.883、0.904,均方根误差分别为1.139、0.271 mg/(100 g)、1.896 mg/(100 g)。结果表明,便携式水产品多品质参数拉曼检测装置可以实现鲳鱼颜色b*、TBA值、TVB-N含量的多指标实时无损检测。

Aquatic products are easy to be corrupted and deteriorated after fishing.However,due to the large volume,inconvenient movement,and complicated operation,it is difficult to be applied for conventional detection devices.In order to meet the needs of rapid detection of aquatic products,a portable multi-quality parameter detection device for aquatic products based on Raman spectroscopy was developed.The hardware system of the device was mainly composed of a light source module,a spectrum acquisition module,a system control processing module,a touch screen,a power module,and a communication module.The Raman spectral data of samples collected by the hardware system of the device,and the collected spectral data was smoothed and background subtracted by standard normal variate transformation,Whittaker smoothing algorithm,adaptive iterative reweighted penalty least squares algorithm.At the same time,according to the national standard method,the standard physicochemical values of pomfret color b*,thiobarbituric acid(TBA)and total volatile base nitrogen(TVB-N)were collected,combined with spectral data,partial least squares regression quantitative prediction model for aquatic product quality was established.The results showed that the correlation coefficients of the validation color of b*,TBA and TVB-N were 0.907,0.897 and 0.915,respectively.The root mean square error of the verification set was 1.315,0.104 mg/(100 g)and 2.875 mg/(100 g).The prediction model was implanted into the portable device software system to realize one-key detection of multi-quality parameter indicators of aquatic products.Finally,the accuracy of portable device was verified.The correlation coefficients between the predicted values of b*,TBA and TVB-N of 18 pomfret samples and standard values were 0.927,0.883 and 0.904,and the root mean square errors were 1.139,0.271 mg/(100 g)and 1.896 mg/(100 g),respectively.The results showed that the portable device can realize the multi-index real-time non-destructive detection of pomfret.