基于Sine-SSA-BP算法智能矫直系统工艺参数预测  

Process parameters prediction of intelligent straightening system based on Sine-SSA-BP algorithm

作　　者：南无疆 王礼先 NAN Wu-jiang;WANG Li-xian(Institute of High-end Heavy Machinery and Equipment,Taiyuan University of Science and Technology,Taiyuan 030024,China;Qingdao Hongjie Intelligent Technology Co.,Ltd.,Qingdao 266111,China)

机构地区：[1]太原科技大学高端重型机械装备研究院,山西太原030024 [2]青岛泓捷智能科技有限公司,山东青岛266111

出　　处：《塑性工程学报》2024年第12期81-89,共9页Journal of Plasticity Engineering

摘　　要：为了改善矫直机工艺参数与板材矫直后板形的关系,并提升板材零件的矫直精度和矫直系统的智能化程度,提出基于Sine-SSA-BP算法来构建智能辊式矫直系统工艺参数选取模型。该算法通过Sine混沌映射初始化麻雀位置,并采用Sine-SSA算法优化BP神经网络的权值和阈值来构建模型。同时,使用MATLAB软件进行仿真,仿真结果与实验结果进行对比,研究结果表明基于Sine-SSA-BP算法建立的模型可有效提升矫直工艺参数预测精度,系统预测模型的均方根误差为0.246,平均绝对百分误差为8.1%,拟合度为0.983,且模型具有更好的鲁棒性,可以为生产中矫直工艺参数选取提供指导。To improve the relationship between the straightening machine process parameters and the shape of the plates after straighte-ning,as well as to enhance the straightening accuracy of plate components and the intelligence level of the straightening system,a model for selecting process parameters for intelligent roller straightening systems based on the Sine-SSA-BP algorithm was proposed.The sparrow positions were initialized by using Sine chaotic mapping,and the model was constructed by optimizing the weights and thresholds of the BP neural network with the Sine-SSA algorithm.Simulations were conducted using MATLAB software and the simulation results were com-pared with experimental results.The research results indicate that the model established based on the Sine-SSA-BP algorithm can effec-tively improve the prediction accuracy of straightening process parameters.The root mean square error of the system prediction model is 0.246,the average absolute percentage error is 8.1%,and the goodness of fit is 0.983.Furthermore,the model exhibites better robust-ness,providing guidance for the selection of straightening process parameters in production.

关 键 词：智能辊式矫直系统 Sine混沌映射 麻雀算法 BP神经网络 工艺参数 

分 类 号：TG333.2[金属学及工艺—金属压力加工]