一种辊式矫直智能优化工艺预测模型的研究与应用  

Research and application of an intelligent optimization process prediction model for roller straightening process

作　　者：胡鹰 原嘉辰 吕畅 HU Ying;YUAN Jiachen;L Chang(College of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学计算机科学与技术学院,太原030024

出　　处：《现代制造工程》2024年第8期109-117,共9页Modern Manufacturing Engineering

基　　金：国家自然科学基金资助项目(52275357,52175354)。

摘　　要：针对传统矫直过程中需要依赖人工经验、矫直速度慢和板材良品率低的问题,综合考虑板材矫直过程中板厚、弹性模量、屈服强度和板材塑性率等参数对矫直工艺的影响,以及反向传播(Back Propagation, BP)神经网络容易陷入局部最优值和泛化能力不强等问题,引入蜣螂优化(Dung Beetle Optimizer, DBO)算法,建立了基于蜣螂优化算法优化BP神经网络的矫直智能优化工艺预测模型。使用包含1 000条数据的训练集进行训练,对比BP神经网络预测模型和粒子群算法优化BP预测模型,结果表明,蜣螂优化算法优化BP神经网络预测模型的首尾辊压下量百分比误差分别在0.5%和0.6%以内,总矫直力百分比误差在0.6%以内,该预测模型对于矫直工艺的精确预测有较高的参考价值。In response to the issues of reliance on manual expertise,slow straightening speed,and low yield rate of quality products in traditional straightening processes,a straightening intelligent optimization process prediction model based on the Dung Beetle Optimizer(DBO)algorithm optimized Back Propagation(BP)neural network was proposed.Considering the influences of parameters such as plate thickness,elastic modulus,yield strength,and plastic ratio of the plate during the straightening process,as well as the issues of BP neural networks easily falling into local optima and weak generalization ability,the DBO algorithm was introduced.The model was trained using a training set consisting of 1000 data points.A comparison between the BP neural network prediction model and the particle swarm algorithm optimized BP prediction model was conducted.Results show that the percentage errors of the DBO algorithm optimized BP neural network prediction model for the adjustment amounts of the head and tail rollers are within 0.5%and 0.6%respectively,and the total straightening force percentage error is within 0.6%.The proposed model demonstrates high reference value for accurate prediction of the straightening process.

关 键 词：矫直工艺 蜣螂优化算法 BP神经网络 预测模型 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]