Study On Optimal Setting Of Control Parameters For Plan View Pattern Of Plate

Plan view pattern control technology of medium and thick plate is a very effective means to improve the rate of finished products and economic benefits of enterprises.In this thesis,using ABAQUS finite element software in the study on metal flow law in the process of medium and thick plate rolling,and the accurate prediction model of plan view pattern was obtained by training the simulation results based on the BP neural network.The moving least squares method was used to optimize prediction come tail curve,improve precision of the prediction model and on this basis,the optimization was studied for the shape control parameters,reducing cutting loss,improve yield and to support the steel plate view control during production.The main research contents of this paper are as follows:(1)The 3D rolling model of medium and thick plate was established through finite element software ABAQUS which was redeveloped based on Python language.The plane view pattern control parameters were set as variables to simplify the modeling process and realize the rapid simulation development of the three-dimensional rolling process of medium and thick plate.(2)Simulate the single-pass and multi-pass rolling process of medium and thick plate under different modes,study the influence of different width-broadening ratio,elongation ratio and control parameters on the metal flow of plate head and tail,and analyze the influence law of plane view pattern control parameters on the shape of finished product by extracting 3D deformation grid coordinates.(3)The metal flow curve during rolling of the plate was obtained by analyzing the simulation data and fitting the metal flow at the head and tail of the plate in the form of 6 curves.With the data of finite element simulation results as the BP neural network training and testing samples,the planar shape prediction model was established based on BP neural network,and then determine the neural network structure,learning algorithm,the number of hidden layer,hidden layer nodes,the input and output parameters,get a intelligent planar shape prediction model with high accuracy by training sample data.(4)Combined with the actual production site process procedures,the developed planar shape prediction model was used to predict the shape of the final rolling product.Based on the calculation results of planar shape and the actual feedback,the head and tail shape curves were optimized by the moving least square method to obtain an accurate prediction model.The plan view pattern control model studied in this paper can improve the precision of the head and tail shape control,reduce the cutting loss rate,improve the product yield and economic benefits of the enterprise.