文献类型：期刊文献

英文题名：Pulse vacuum pretreatment technology and neural network optimization in drying of tilapia fillets with heat pump

作者：Li, Min[1,2];Wu, Yangyang[1,2];Ge, Yunting[3];Ling, Changming[1,2]

机构：[1]Guangdong Ocean Univ, Shenzhen Inst, Shenzhen 518108, Guangdong, Peoples R China;[2]Guangdong Ocean Univ, Coll Mech & Power Engn, Zhanjiang, Peoples R China;[3]Univ South Wales, Sustainable Environm Res Ctr, Pontypridd CF37 1DL, M Glam, Wales

年份：2019

卷号：43

期号：12

外文期刊名：JOURNAL OF FOOD PROCESSING AND PRESERVATION

收录：SCI-EXPANDED(收录号：WOS:000491721500001)、、EI(收录号：20194507624177)、Scopus(收录号：2-s2.0-85074339097)、WOS

基金：Natural Science Foundation of Guangdong Province, Grant/Award Number: 2015A030313613; Science and Technology Department Project of Guangdong Province, Grant/Award Number: 2014A020208115; Shenzhen Science and Technology Innovation Committee, Grant/Award Number: JCYJ20170306162116012; Guangdong Science and Technology Department, Grant/Award Number: 2017A010104011

语种：英文

外文关键词：Neural networks - Pumps - Drying - Parameter estimation - Heat pump systems - Quality assurance

外文摘要：Low-temperature (20 similar to 40 degrees C) heat pump drying technology for food products has advantages over conventional high temperature (45 similar to 90 degrees C) hot air drying but requires proper pretreatment to ensure high-quality drying products. Subsequently, experimental investigations were carried out for drying of tilapia fillets with heat pump and at different pulse vacuum pretreatment parameters including vacuum cycle rate, cycle number, vacuum pressure, and concentration of trehalose impregnation liquid. The quality of the completed drying product was then evaluated and compared by eight quality indexes containing whiteness, ratio of water loss to solid content increase rate during osmosis dehydration (dehydration efficiency index, DEI), Ca2+-ATPase (adenosine triphosphatase) activity, rehydration rate, hardness, elasticity, glueyness, and chewiness. These indexes are thus generalized into one comprehensive score based on the effective weight of each index on the product quality and analysis of the fuzzy hierarchy process. The experimental measurements were applied to examine the effect of the pretreatment parameters on the product quality indexes and comprehensive score. Accordingly, an artificial neural network model has been developed to train and achieve the optimal pretreatment parameters based on the maximum product quality comprehensive score. The obtained optimal pretreatment parameters have been validated with the measurements. Ultimately, the optimal heat pump drying process is significant and can be applied to an actual application for different food product dryings. Practical applications Heat pump technology is one of the prospective processing methods for the drying of tilapia fillets. However, the relative low-temperature heat pump drying process would increase the drying time and thus deteriorates the risk of food product. To cope with this, a high efficient pretreatment method needs to be designed and applied. Correspondingly, in the project described by this paper, extensive measurements were designed and carried out on the product drying with a heat pump. Consequently, a fuzzy hierarchy process was applied to process the relevant experimental data and generate a comprehensive score for each test. Ultimately, an artificial neural network model has been developed to train and achieve the optimal pretreatment parameters. The optimal heat pump drying process is significant and can be applied to an actual application for different food product dryings.