>> 自然科学版期刊 >> 2021 >> 2021年05期 >> 正文
基于MEA-BP神经网络的超声挤压加工表面粗糙度预测
时间: 2021-09-10 次数:

陈爽, 张志, 肖锦初, .基于MEA-BP神经网络的超声挤压加工表面粗糙度预测[J].河南理工大学学报(自然科学版),2021,40(5):104-109.

CHEN S, ZHANG Z, XIAO J C, et al.Surface roughness prediction of ultrasonic extrusion processing based on MEA-BP neural network[J].Journal of Henan Polytechnic University(Natural Science) ,2021,40(5):104-109.

基于MEA-BP神经网络的超声挤压加工表面粗糙度预测

陈爽, 张志, 肖锦初, 胡家进, 赵录冬

江西理工大学 机电工程学院,江西 赣州341000

摘要:为了有效预测超声挤压加工工件的表面粗糙度,建立以转速,进给速度,振幅,挤压力,挤压次数为输入参数,表面粗糙度为输出结果的预测模型。该模型利用思维进化算法(mind evolutionary algorithm MEA)的全局搜索能力对BP神经网络的权值和阈值进行优化。为了验证该模型的有效性,对45号钢进行超声挤压加工后,使用BP神经网络进行预测,通过引入思维进化算法(MEA)和遗传算法(GA)优化BP神经网络的权值和阈值,并对3种模型的预测精度进行对比分析。结果表明:在相同的实验条件下,MEA-BP模型的预测结果最精准,与BP神经网络相比,该模型精度高,运行速度快。

关键词:超声挤压加工;表面粗糙度预测;思维进化算法;BP神经网络;预测精度

doi:10.16186/j.cnki.1673-9787.2020060052

基金项目:国家自然科学基金资助项目(51965023);江西省高校科技落地计划项目(KJLD14044);江西省自然科学基金资助项目 20151BBE50037

收稿日期:2020/06/17

修回日期:2020/09/15

出版日期:2021/09/15

Surface roughness prediction of ultrasonic extrusion processing based on MEA-BP neural network

CHEN Shuang, ZHANG Zhi, XIAO Jinchu, HU Jiajin, ZHAO Ludong

College of Mechanical and Electrical Engineering Jiangxi University of Science and Technology Ganzhou 334000JiangxiChina

Abstract:In order to effectively predict the surface roughness of the workpiece after ultrasonic extrusion processing a prediction model was established with rotational speed feed speed amplitude extrusion pressure and extrusion times as input parameters and surface roughness as output parameters. The weights and thresholds of BP neural network were optimized in this model by using the global search ability of mind evolutionary algorithmMEA. In order to demonstrate the effectiveness of the model BP neural network was used to predict 45steel after ultrasonic extrusion processing. The weights and thresholds of BP neural network were optimized by introducing mind evolutionary algorithmMEA and genetic algorithmGA and the prediction accuracy of the three models was compared and analyzed. The results showed that the prediction model of MEA-BP was the most accurate under the same experimental conditions. Compared with BP neural network the proposed model had higher accuracy and faster running speed.

Key words:ultrasonic extrusion processing;surface roughness prediction;mind evolutionary algorithm;BP neural network;prediction accuracy

 基于MEA-BP神经网络的超声挤压加工表面粗糙度预测_陈爽.pdf

最近更新