Modeling And Simulation Of Polyacrylonitrile-based Carbon Fiber Solidification And Pre-oxidation Process

After years of development,carbon fiber has become a new type of material with a wide range of applications.Polyacrylonitrile fiber is a material for preparing carbon fiber,and has important applications in the field of carbon fiber preparation.After years of catching up,my country’s carbon fiber production process is currently at the world’s advanced level,but the large-scale production capacity of high-end carbon fiber still needs to be improved.The carbon fiber preparation process is complex.This paper combines mathematical modeling with 3D animation technology,uses machine learning algorithms to establish a mathematical model of the carbon fiber production process,predicts product performance indicators,and embeds it into a 3D visual dynamic model.Visual display can not only help beginners better understand the process flow,play a role in training and display,but also provide a reference for the optimization of process conditions in the production process,which has important application value to the fine carbon fiber production process.In this paper,the extreme learning machine(ELM)model is used to model the experimental data of the solidification forming and pre-oxidation process,and the prediction results of the model output can be obtained by inputting the process condition parameters in the model;the multi-core extreme learning machine(MKELM)model is introduced to solve the problem.The problem of unstable prediction results caused by random generation of extreme learning machine parameters has been improved,and the stability of the extreme learning machine prediction model has been improved;the genetic algorithm(GA)is used to optimize the parameters of the multi-core extreme learning machine,and the GA-MKELM prediction model is established to solve The problem that the MKELM model is easy to fall into local optimization is solved,and the prediction accuracy and stability are further improved.In this paper,Unity3 D is used to visualize the process flow of solidification and pre-oxidation,and the GA-MKELM prediction model is embedded into the visualization model.The simulation model shows the process and parameter changes of pre-oxidation and solidification respectively.The model is combined with the visual model,and the parameter prediction function is realized in the visual model.Combining the latest animation production technology with chemical process visualization,and using the current rapidly developing data modeling method to simulate and model experimental data such as process parameters can better realize the practical application of the technology.The combination of the three will help It can greatly improve the work efficiency of technicians,and can use software for reference in actual operation to optimize data models.In the field of teaching and popular science,it can increase the interest of beginners,and deepen the understanding of chemical technology,so as to achieve multiple purposes.