光皮木瓜真空脉动干燥特性及神经网络模型  

Drying characteristics of Chaenomeles sinensis with vacuum pulsed drying technology based on BP neural network mode

作　　者：巨浩羽 赵士豪[1] 赵海燕 张卫鹏 肖红伟[4] JU Hao-yu;ZHAO Shi-hao;ZHAO Hai-yan;ZHANG Wei-peng;XIAO Hong-wei(College of Bioscience and Engineering,Hebei University of Economics and Business,Shijiazhuang,Hebei 050061,China;College of Business Administration,Hebei University of Economics and Business,Shijiazhuang,Hebei 050061,China;College of Artificial Intelligence,Beijing Technology and Business University,Beijing 100048,China;College of Engineering,China Agricultural University,Beijing 100083,China)

机构地区：[1]河北经贸大学生物科学与工程学院,河北石家庄050061 [2]河北经贸大学工商管理学院,河北石家庄050061 [3]北京工商大学人工智能学院,北京100048 [4]中国农业大学工学院,北京100083

出　　处：《食品与机械》2022年第3期147-153,共7页Food and Machinery

基　　金：国家自然科学基金项目(编号:32102141);河北省自然科学基金资助项目(编号:C2020207004);河北省高等学校科学技术研究项目(编号:QN2021054);北京市自然科学基金项目(编号:6204035)。

摘　　要：目的:提高光皮木瓜干燥效率和品质。方法:以光皮木瓜为试验原料,选取干燥温度(50,60,70℃)、真空时间(5,10,15,20 min)和常压时间(2,4,8 min)为影响因素,以干燥时间、复水比、维生素C含量、总黄酮含量、微观结构为指标进行单因素试验,建立BP神经网络模型并验证模型的预测效果。结果:干燥温度、真空时间和常压时间对光皮木瓜的干燥时间影响显著(P<0.05),其有效水分扩散系数为6.0448×10^(-10)~12.0086×10^(-10) m^(2)/s,且D_(eff)随干燥温度的升高而增大。BP神经网络模型由干燥时间、干燥温度、常压时间和真空时间4个输入神经元、7个隐含层和含水率1个输出神经元构成。当干燥温度为65℃、常压时间为3 min、真空时间为12 min时,模型的预测值和实测值最大误差为4.77%。光皮木瓜的复水性随干燥温度的提高而降低,随常压时间和真空时间的延长而先升高后降低;维生素C、总黄酮含量随干燥温度、常压时间和真空时间的增加先上升后降低。当干燥温度为70℃时,物料表面因大量失水而导致结壳硬化,水分迁移孔道坍塌堵塞;当干燥温度为50℃时,物料表面呈蜂窝状多孔结构,有助于水分扩散迁移。结论:真空脉动干燥光皮木瓜的最佳工艺条件为干燥温度60℃、真空时间10 min、常压时间4 min,该条件下干燥时间12.1 h、复水比6.28±0.05、维生素C、总黄酮含量分别为(71.26±0.74)×10^(-2),(19.27±0.33)mg/g,BP神经网络模型可以很好地描述光皮木瓜的真空脉动干燥过程。Objective:This study aimed to investigate the drying characteristic of Chaenomeles sinensis by using vacuum pulsed drying technology and establish BP neural network model.Methods:The single factor experiment of drying temperature(50,60,70℃),constant atmosphere time(2,4,8 min)and vacuum time(5,10,15,20 min)on drying time,rehydration ratio,V_(C) and general flavone content as well as microstructure of Chaenomeles sinensis during vacuum pulsed drying technology were investigated.Results:All the drying temperature,constant atmosphere time and vacuum time had significant influence on drying time(P<0.05).The moisture effective diffusion coefficient(D_(eff))ranged from 6.0448×10^(-10) to 12.0086×10^(-10) m^(2)/s in different drying conditions and increased with drying temperature increasing.BP neural network mode consisted of input layer,hidden layer and output layer.The input layer included four neurons named drying time,drying temperature,constant atmosphere time and vacuum time.The hidden layer included seven neurons and the output layer included one neuron named moisture content.The maximum error between simulated and experimental values was 4.77%.Rehydration ratio decreased as drying temperature increased and increased first and then decreased with the extension of atmospheric pressure time and vacuum time.V C and general flavone content increased first and then decreased with the increasing of drying temperature,atmospheric pressure time and vacuum time.The microstructure indicated that when drying temperature was 70℃,the material surface crusted due to a large amount of water loss.In this case,the water migration channel collapsed and blocked.When drying temperature was 50℃,the surface of the material appeared a cellular porous structure,which was conducive to water diffusion and migration.Conclusion:The optimal drying process was drying at temperature 60℃,with atmospheric pressure for 5 min and vacuum for 15 min.In this circumstance,the drying time,rehydration ratio,V_(C) and general flavone content we

关 键 词：光皮木瓜 真空脉动干燥 BP神经网络 复水比 维生素C 总黄酮 

分 类 号：TS255.1[轻工技术与工程—农产品加工及贮藏工程] TP183[轻工技术与工程—食品科学与工程]