Research On Link Prediction Of Konjac Glucomannan Gel Network And Its Properties

Recently,in the wake of the rapid expansion of science and technology,computer is widely used in various disciplines and fields.In the research process,People often encounter the phenomenon that experiments are not easy to explain or experimental methods are not easy measure the amount,then we need use the computer theory to study these experimental phenomena and calculate unknown quantity.Furthest more,the computer can apply prediction and simulation to verify the accuracy of the experiment,and to conduct the experiment.At last,research must base on experiment,and combine with the theory to speed up the progress of research.KGM is the main component of konjac tuber,as a good source of soluble dietary fiber,has a unique mechanical properties and excellent healthy functions.As a raw material or additive,it has been increasingly widely used in the food industry.Due to its structure is complex,the current experimental methods are also difficult to do in-depth study of its structure and properties.There are many report about studies on its structure,but its exact primary structure have no unified conclusion.Therefore,we try applying network theory with thoughts of the computer to study KGM.First of all,we use the experiment to study the its primary structure,and combine with link prediction to evaluate the accuracy of the experiment result.Then,we use finite element method to study the relationship between a series of network parameters and KGM gel network.The details are specified as follows:1.Determination of the KGM structureBy combining instrumental analysis with the references,it is reasonable to assume its primary structure that:KGM is heteropolysaccharide that composed of glucose and mannose in a 1:1.79 molar ratio that connected by β-1,4 glycosidic bond.In the main chain,about every sixty-eight sugar residues connect a branched chain.The length of branched chain is three-four sugar residues that connected by β-1,3 glycosidic bond in C3 position of mannose,and each of nineteen sugar residues on the backbone connect a acetyl acetate2.The Link Prediction of KGMWe use four kinds of similarity index to do link prediction on KGM network,and then use the index AUC to evaluate link prediction results,and combining with the results of instrumental analysis determined the structure,there were made in the following conclusions:the prediction results of local path index-LP are optimal,and the minimum of four kinds index of AUC values are more than 0.5,and most AUC value are about 0.8,and even the maximum of AUC value can reach 0.9;it was combined result of instrument analysis with the result of link prediction and shown that the primary structure prediction KGM is quite accurate.3.The influences of the network characteristic parameters on its gel mechanical propertiesRecently,although people do many researches on the type of the gel,mechanism,mechanical properties,and the three-dimensional network structure.But research on the relationship of the mechanical properties and its gel structure has not been reported.we analyze,understand the experimental results based on the available experiment,and establish a three-dimensional network model of KGM gels.Then,we observed the morphology of different concentration of KGM water solution by AFM and obtained the best concentration that establishs the best network model.And last we use the finite element analysis simulation software-ANSYS to simulate the mechanical properties of KGM gel network and study the influence of the network parameters on the mechanical properties of the network.From the observation and simulation results and analysis of the results the following conclusions can be drawn:(1)When the KGM concentration was 60g/ml,KGM to form a three-dimensional network structure of mutual winding,this concentration structure of konjac glucomannan gel network model is the best.(2)Increasing the number of nodes of the network can make the network elastic modulus-G increase.(3)By comparing the model and the actual structure of KGM organic gel network we have come to conclusion:when the number of small network is largely unchanged,reduce the length between the nodes of the network-l to result in the network elastic modulus-G dramatically increasing;when the length between the nodes of the network-l is roughly constant,reduce the number of small network to make the network elastic modulus-G increase.(4)When the ratio of the diameter between the nodes of the network-l and the network radius of cross section-r is less than twenty,and increasing the radius of cross section of the network-r can make the network elastic modulus-G increase;when the ratio of the diameter between the nodes of the network-l and the network radius of cross section-r is more than twenty,the radius of cross section of the network-r has almost no effect on the network elastic modulus-G.