Vacuum freeze-drying technology (referred to as ‘freeze-drying’) can effectively prevent the decomposition of heat-sensitive substances and retain the nutrition, color and form of foodstuffs to the maximum extent.However, it is difficult to accurately determine the end point of each stage of freeze-drying, and it is common in the market to extend the drying time to ensure the degree of drying, which often leads to a decline in the quality of the final product.The paper aimed to optimize the vacuum freeze-drying process and improve the product quality using dragon fruit as an example.Through orthogonal experimental design, this paper adopted the polar analysis method to explore the influence of key process parameters such as freezing temperature, vacuum degree, sublimation temperature and resolution temperature on the sensory quality of the product, and carried out validation experiments based on the best combination of sensory quality process parameters obtained,compared and analysed the fitting effects of the BP neural network, support vector regression model and random forest regression model on the data of the orthogonal experiments, and the models required a large amount of sample data for training, therefore, it was not possible to analyze the effect of the models on the data of the orthogonal experiments.The virtual samples were added in this study because the models require a large amount of sample data for training.The results showed that the order of influence on sensory quality was resolution temperature＞ dragon fruit thickness＞ vacuum＞ freezing temperature＞ sublimation temperature.The BP neural network had the best fitting effect.It was verified that the freezing time of dragon fruit was 36 min, the vacuum time was 11 min, the sublimation time was 784 min, the resolution time was 111 min, and the quality score was 14.12, and the predicted values of the five indexes were all were close to the measured values with the highest relative error of 6.57%.
[1] 罗瑞明,周光宏,乔晓玲.干切牛肉冷冻干燥中高速率升华条件的动态研究[J].农业工程学报, 2008, 24(2):226-231.
LUO R M,ZHOU G H,QIAO X L.Dynamic study on operation conditions for the higher sublimation rate during freezing drying of cooked beef slice[J].Transactions of the Chinese Society of Agricultural Engineering, 2008, 24(2):226-231.
[2] 张彤,余克志,张得正.火龙果真空冷冻干燥的模拟分析和实验研究[J].制冷学报,2022,43(2):142-150.
ZHANG T, YU K Z, ZHANG D Z.Simulation analysis and experimental research on vacuum freeze-drying of pitaya[J].Journal of Refrigeration,2022,43(2):142-150.
[3] 姜秀丽,牛海力,孔保华,等.烘干温度对牛肉干水分分布与品质相关性的影响[J].食品研究与开发,2016,37(21):10-15.
JIANG X L, NIU H L, KONG B H, et al.The relationship between water distribution and quality of beef jerky as attributed to different drying temperatures[J].Food Research and Development,2016,37(21):10-15.
[4] 孟繁博,黄道梅,郑秀艳,等.超声波预处理对热风干燥火龙果片品质的影响[J].食品与发酵工业,2021,47(13):205-209.
MENG F B, HUANG D M, ZHENG X Y, et al.Ultrasound pretreatment on quality of pitaya before hot air drying[J].Food and Fermentation Industries,2021,47(13):205-209.
[5] 齐建东,黄金泽,贾昕.基于XGBoost-ANN的城市绿地净碳交换模拟与特征响应[J].农业机械学报,2019,50(5):269-278.
QI J D, HUANG J Z, JIA X.Simulation of NEE and characterization of urban green-land ecosystem responses to climatic controls based on XGBoost-ANN[J].Transactions of the Chinese Society for Agricultural Machinery,2019,50(5):269-278.
[6] 向鹏成,高天,段旭,等.基于机器学习的“一带一路”投资国别风险预测研究[J].工业技术经济,2024,43(7):150-160.
XIANG P C, GAO T, DUAN X, et al.Investment country risk prediction of “the Belt and Road” based on machine learning[J].Journal of Industrial Technology and Economy,2024,43(7):150-160.
[7] 王韦燕,冯文强,常乃杰,等.基于光谱预处理和机器学习算法的烤烟叶绿素含量预测[J].中国土壤与肥料,2023(3):194-201.
WANG W Y, FENG W Q, CHANG N J, et al.Prediction of chlorophyll content in flue-cured tobacco based on spectral pretreatment and machine learning algorithm[J].Soil and Fertilizer Sciences in China,2023(3):194-201.
[8] 孙可欣,范蕊,郑彬,等.基于机器学习算法的能源总线系统数据驱动建模研究[J].建筑科学,2024,40(4):248-259.
SUN K X, FAN R, ZHENG B, et al.Study on data-driven modeling of energy bus system based on machine learning algorithms[J].Building Science,2024,40(4):248-259.
[9] 加世滢,王清华,王贞艳,等.无缝钢管斜轧穿孔管形多目标预测[J].锻压技术,2022,47(10):169-175.
JIA S Y, WANG Q H, WANG Z Y, et al.Multi-objective prediction on cross-rolling and perforated pipe shape for seamless steel pipe[J].Forging & Stamping Technology,2022,47(10):169-175.
[10] 刘燕君,刘凯,曹晶晶.结合珠海一号高光谱影像和XGBoost算法的珠江口滨海湿地分类[J].测绘通报,2023(12):136-141.
LIU Y J, LIU K, CAO J J.Classification of coastal wetlands in the Pearl River Estuary using Zhuhai-1 hyperspectral imagery and XGBoost algorithm[J].Bulletin of Surveying and Mapping,2023(12):136-141.
[11] BOSS E A, FILHO R M, DE TOLEDO E C V.Freeze drying process:Real time model and optimization[J].Chemical Engineering and Processing:Process Intensification, 2004, 43(12):1475-1485.
[12] 胡国田,尚会威,谭瑞虹,等.不同土壤类型的有机质含量的可见-近红外光谱检测模型传递方法研究[J].光谱学与光谱分析,2022,42(10):3148-3154.
HU G T, SHANG H W, TAN R H, et al.Research on model transfer method of organic matter content estimation of different soils using VNIR spectroscopy[J].Spectroscopy and Spectral Analysis,2022,42(10):3148-3154.
[13] 乔宏柱,高振江,王军,等.大蒜真空脉动干燥工艺参数优化[J].农业工程学报,2018,34(5):256-263.
QIAO H Z, GAO Z J, WANG J, et al.Optimization of vacuum pulsed drying process of garlic[J].Transactions of the Chinese Society of Agricultural Engineering,2018,34(5):256-263.
[14] 马诗瑜,何敬成,詹陆川,等.基于人工神经网络算法的自拟清瘟方制备工艺优化探索[J].中国药业,2023,32(12):56-62.
MA S Y, HE J C, ZHAN L C, et al.Optimization of preparation process of self-prepared Qingwen formula based on artificial neural network[J].China Pharmaceuticals,2023,32(12):56-62.
[15] 袁蕉蕉,赵武奇,于维平,等.真空冷冻干燥猕猴桃果粉的工艺优化[J].真空科学与技术学报,2021,41(1):69-74.
YUAN J J, ZHAO W Q, YU W P, et al.Optimizationof vacuum freeze-drying conditions of kiwifruit powder[J].Chinese Journal of Vacuum Science and Technology,2021,41(1):69-74.
