基于NIR的鲜食葡萄物流过程安全系统

针对鲜食葡萄物流过程中的实际模式及通常会使用SO₂保鲜剂的情况，研究了4种温度及定量SO₂胁迫条件下，基于近红外光谱和质构变化的鲜食葡萄货架期预测方法，结合信息技术设计了基于货架期预测的物流过程安全系统，以期减少鲜食葡萄物流过程中的损失。质构变化是鲜食葡萄采后到达货架期终点的重要原因，研究使用SO₂浓度传感器控制电磁阀，通过SO₂自动补偿获得设定SO₂浓度，研究了不同浓度SO₂胁迫条件下鲜食葡萄的质构变化规律及温度的影响。对比了多元散射校正和一阶S-G求导预处理方法对原始光谱的预处理效果，采用偏最小二乘回归方法建立了基于近红外光谱的鲜食葡萄质构无损检测模型，模型决定系数为0.93，均方根误差为1.70，通过交叉验证，模型预测准确度为0.81，均方根误差为2.91。研究表明，近红外快速无损检测可结合品质变化建模和信息技术用于提高果蔬采后物流过程安全管理效率。

Logistics Process Safety System of Table Grapes Based on NIR

In view of the fact that fresh keeping agents based on sulfur dioxide are commonly used in table grape logistics and actual logistics process of table grapes,we studied the shelf life prediction method of table grapes under 4 temperatures and constant concentrations of sulfur dioxide based on near infrared spectrum (NIR)and the evolution of texture in this work. According to multi-stage logistics process of table grapes,logistics process safety system based on cloud computing was designed to reduce the loss of table grapes in the logistics.The change of texture is an important cause of table grapes to end their shelf life in postharvest logistics with the using of sulfur dioxide.The technology of NIR has advantages of rapid and nondestructive testing.It would be of great realistic significance as to the combination of NIR,food quality dynamic change and information technology which are to be used to the logistics process safety management.In this work,we used SO2 concentration sensors to control solenoid valves,and obtained the set SO2 concentrations by automatic compensation mechanism.The evolutions of table grape texture under different concentrations of sulfur dioxide were studied as well as the influence of temperature.The NIR pre-treatment effects of multiplicative scatter correction and the first S-G derivation were compared.The table grape texture nonde-structive testing model built base on NIR and partial least squares regression achieved a determination coefficient of 0.93 and the root mean squared error (RMSE)was 1.70.In full cross-validation,the prediction accuracy reached to 0.81 and got a RMSE of 2.9 1 .The consistency of processing results of local software and cloud computing was verified.And the comparative test showed that the same processing results could be gained from both methods.The combined applications of NIR nondestructive testing and information technology to the postharvest logistics management of fruits and vegetables could realize the left shelf life early warning based on rapid detection and continuous monitoring and improve the precise of logistics management and food security.