The Application Of BP Neural Network In Cellulose Nanofibers

As a green polymer material,cellulose is a renewable resource.Its wide range of sources and degradability are consistent with the current theme of green development.Electrospinning can directly and continuously prepare polymer nanofibers.Due to its large specific surface area,controllable morphology and diameter,and outstanding functionality,the nanofibers are widely used in high-efficiency filter materials,flexible wearable sensors,supercapacitors,and so on.There are a wide range of applications for functional products such as luminescence and conductive,especially.If the advantages of both can be combined,the cellulose nanofibers are prepared by electrospinning technology,so that the fibers retain the excellent properties of cellulose degradation,and at the same time have the function of nanofibers,then the prepared products will have important application value.However,direct electrospinning of cellulose to prepare nanofibers faces great challenges.In this paper,the BP neural network model is first combined with the electrospinning process.Based on the experimental data of 17 kinds of polymers such as polyester and the corresponding 100 sets of polymer solution system and the actual electrospinning results,a theoretical model for predicting the results of polymer electrospinning was established.Among them,polymer structural parameters(solubility parameters),polymer solution characteristic parameters(viscosity,surface tension,conductivity)were used as input.The accuracy of the theoretical model reached 81%through the verification of the remaining 65 sets of polymer spinning solution systems.In order to verify the practicability of the model,three kinds of polymers such as polyethylene oxide and the corresponding 12 groups of polymer solution systems were used.Secondly,the theoretical model of BP neural network theoretically simulate the characteristic parameters and structural parameters of cellulose electrospinning solution.Through theoretical model prediction of spinning results and testing and characterization of experimental results,the most suitable cellulose type for electrospinning was found,and the optimal polymerization degree(DP)was 650.Through the rheological property test of the cellulose electrospinning solution,when the mass percentage is between 2.6%and 2.9%,and the spinning effect is the best.and the optimum spinning temperature was explored at 65 ?.According to the theoretical simulation and rheological properties of cellulose electrospinning solution,DMAc/LiCl was selected as the solvent system to dissolve cellulose,and nanofibers were prepared by electrospinning technology.SEM photographs of the prepared samples showed that the prepared fibers had a smooth surface and the fibers had an average diameter of 400 nm.The TG and tensile test results show that the prepared cellulose nanofibers have good heat resistance and mechanical properties.