Study On The Influence Of Biodiesel Composition And Structural Constituent On Its Key Quality

For the purpose of slowing the consumption of fossil fuels and environmental protection,biodiesel has received worldwide attention due to its advantages of renewable raw materials,mature production processes,and low pollution from combustion emissions.Biodiesel is mainly composed of fatty acid methyl esters(FAME).Currently,it is mainly used for diesel engines and heating,in order to improve the combustion performance of biodiesel and improve emissions,and because of the composition and molecular structure of fatty acid methyl esters,it has low temperature fluidity and oxidation stability.The impact of this is a mutual system.This article selects the low temperature fluidity and oxidation stability as the two key qualities discussed in this paper.In the actual production process,it takes time and labor to measure the physical and chemical properties of a large number of oil products.Therefore,it is of great significance to find a model that can better predict the physical and chemical properties of biodiesel.Based on the components of biodiesel,this paper uses BP neural network to predict the low-temperature fluidity and oxidation stability of biodiesel,and studies the influence of biodiesel composition and molecular structure on its two key qualities,which can be used to predict Ternary phase diagram of biodiesel performance.The back propagation(BP)network was a front feedback neural network,which was consisting of input layer,hidden layer and output layer.Its transmission functions were nonlinear,and the most common functions were the logarithm S type(logsig)function and the hyperbolic tangent S type(tansig)function.The learning method belongs to supervision study.One of the most widely used fields of BP neural network was prediction.Based on BP neural network,this article predicted the physical and chemical properties of biodiesels by using its components.The BP network model contains one input layer,three hidden layers and one output layer.At each training,the number of neuron nodes in the input layer was 23,the number of neuron nodes in the output layer was 1,and the number of neurons in the hidden layer was 47,47,32,32,40,and 45,respectively.The results showed that the error between the actual and predicted values of CFPP,viscosity and induction period was about 2%,3%and 1.5%,respectively.The high content of SFAME in biodiesel tends to have relatively low CFPP by analyzing the effect of FAME composition and structure on the cold flow property of biodiesel.The long straight chain of SFAME tends to have relatively.structural difference with UFAME molecules,which leading crystallize easily.In addition,as the number of double bonds of carbon chain in UFAME increases,the degree of bending of the carbon chain increases,leading steric resistance increase during the migration of the crystalline molecules.So,the intermolecular force was weakened and the crystallization of biodiesel was formed difficulty.The Cold flow property of biodiesel will be better.The biodiesel could be considered as pseudo-binary solution using high melting point as solute and low melting point as solvent The high degree of similarity between the solute and the solvent tend to have relatively great solubility of the solute,and the biodiesel will not be crystallized easily.The influencing factors of oxidation stability of biodiesel were complex.After analyzed-the.effect of FAME composition and structure on the induction period of biodiesel,it was found that the higher content of biodiesel SFAME will leads the longer induction period.The higher content of linoleic acid(C18:2)and linolenic acid/eoispic acid(C18:3)in biodiesel will leads the worse stability.Finally,based on the influence of biodiesel composition and molecular structure on its quality,a ternary phase diagram with SFAME,MUFAME,and PUFAME as three factors was drawn and the corresponding regions were delineated to screen high quality biodiesel.