Relationships Between Thermo-physical Properties And Physiological And Biochemical Parameters Of Fruits And Vegetables And Neural Network Prediction Model On Thermo-physical Properties

With the development of the social and the improvement of people's living standards, People growing demand fresh, nutrition and health food. Therefore, the preservation of fresh food needs a higher requirement, In the course of storage, processing, transport of the fresh foods; the calories transfer and exchanging exist all the time. So thermo-physical properties should be used to choose process craft and equipment design. the parameters of thermo-physical properties is the right important index of Cold Storage and design parameters of refrigeration equipment, but also the right important basis for refrigeration, freezing and dry processing time . At home and abroad there have been many studies on thermo-physical properties, but the research on the inherent relationship between thermo-physical properties of fruits and vegetables and physiological and biochemical parameters is less, and the various parameters of food thermo-physical properties given in the literature is very limited, The thermo-physical properties of food (including thermal conductivity, heat capacity and freezing, etc.) not only with temperature, but also with the relevant of food physical and chemical properties(including moisture content, density, soluble solids content). In view of this, it is necessary to study on this topic in-depth.Artificial neural network which is a kind of analysis method extensive attention in recent years can solve complex nonlinear problems, and overcome non-adaptive of the statistical empirical methods to a greater extent, good at extracting and analyzing macro-statistical law from a large number of statistical, whose main characteristic is non-linear mapping ability, this ability enable them to be well for any non-linear function approximation, and thus obtain more accurate prediction models.The fresh fruit and vegetable considered in this work are bought form the market of shanghai. In this paper through measuring the thermal conductivity, heat capacity and freezing point of many kinds of fruits and vegetables, discusses the relationship between thermo-physical properties and physiological and biochemical parameters including density, total soluble solids content and water content at different states. For different kinds of fresh-ripe fruits and vegetables, we select red delicious, dragon fruit, banana, yacon, radish, tomato, luffa etc as the research objects. The thermo-physical properties of fruits and vegetables (including thermal conductivity, heat capacity and freezing, etc.) and physiological and biochemical parameters(including density, water content and soluble solids content) are measured in accordance with the order of measurement. When measuring the heat capacity, freezing point, water content and soluble solids content,the samples are taken from central of the pulp fruit and measured three times and averaged. In order to find factors which mostly affect thermo-physical properties of fruit-vegetable structure , analyze the data by calculator base on large quantity of tested data, and confirm the connection between thermo-physical properties and index of physiological and biochemical parameters and found the predicting equation of thermo-physical properties of fruit-vegetable. The results show that thermal conductivity of fruits and vegetables increase with temperature, but not very obvious. it is positive correlation with water content and density; negative correlation with soluble solids content, and find the single factor regress equation between thermal conductivity and water content/density/ soluble solids content, simultaneity,find the forecast equation of influence to thermal conductivity by water content/density/ soluble solids content at 5℃, 15℃, 25℃. The forecast equation at 5℃, 15℃, 25℃separately isλ=-0 .5152+0.5771d +0.0044w+0.0014sλ=-0 .4886+0.5558d +0.0045w+0.0016sλ=-0 .5275+0.5198d +0.0053w+0.0025sThe freezing point of fruits and vegetables increase progressively with the water content, decrease progressively with the soluble solid content, and has no obvious correlation with the density, and find the single factor regress equation between freezing point and water content/soluble solids content, simultaneity,find the forecast equation of influence to freezing point by water content/soluble solids content. The forecast equation is T = ?4.8755+0.0429w ?0.3379s.The heat capacity of fruits and vegetables increase progressively with the temperature, and as the temperature increases, the growth rate of the heat capacity of fruits and vegetables decreases until no change, There is a high positive correlation between heat capacity and water content, negative correlation between heat capacity and soluble solids content, and has no obvious correlation between heat capacity and the density, and find the single factor regress equation between heat capacity and water content/soluble solids content, simultaneity,find the forecast equation of influence to heat capacity by water content/soluble solids content at 10℃, 15℃, 20℃. The forecast equation at 10℃, 15℃, 20℃separately is Cp = 1 ..3070+0.0219w-0.0096s; Cp = 1 .2475+0.0219w?0.0089s; Cp = 1 .1674+0.0253w?0.0061s.Finally, the paper briefly describes BP artificial neural network and the creation of artificial neural networks. A BP artificial neural network model was presented for the prediction of thermo-physical properties of fruits and vegetables as a function of temperature, water content, density ,soluble solid content and other parameters. After the comparative analysis of the error, the optimal prediction model can be established, and then compared to the actual test results, verify the reliability of the model. The results show that the model used for the thermo-physical properties of fruits and vegetables have good prediction accuracy and can be applied to the prediction of thermo-physical properties of fruits and vegetables. In practice, due to temperature, the water content, density, soluble solids content and other parameters easily measured, thermo-physical properties values of fruits and vegetables can be predicted according to the optimized BP neural network model and the predicted speed, simple operation, which can provide the reliable theory and data for Food storage processing industry .