您的位置:首页 > 论文页面
超弹性形状记忆合金循环特性的BP神经网络模型研究
发表时间:2011-07-31 浏览量:1548 下载量:428
全部作者: | 任文杰,贾茹,韩雷 |
作者单位: | 河北工业大学土木工程学院 |
摘 要: | 对超弹性形状记忆合金(shape memory alloy,SMA)丝进行循环加卸载试验,在此试验基础上,采用误差反向传播(back-propagation,BP)神经网络原理,建立超弹性SMA的循环本构模型。该模型的输入为循环次数、加卸载信息和应变值,输出为应力值。从试验曲线中选择25条用于训练网络,5条用于检验网络。计算证明了该模型具有很高的预测精度。这为超弹性SMA循环本构模型的建立提供了一个新的思路。 |
关 键 词: | 防灾减灾工程及防护工程;形状记忆合金;超弹性;倒传递类神经网络;本构模型 |
Title: | Study of BP neural network model of cyclic behavior of super-elastic shape memory alloy |
Author: | REN Wenjie, JIA Ru, HAN Lei |
Organization: | School of Civil Engineering, Hebei University of Technology |
Abstract: | In this study, the cyclic tensile tests were carried out on the super-elastic shape memory alloy (SMA) wires. According to the testing data, a constitutive model for the cyclic behavior of super�elastic SMA was put forward based on back�propagation (BP) neural network. In this model, the inputs included the number of loading cycles, the index of loading and unloading and the strain; and the output was the stress. 25 experimental curves were used for training the neural network and 5 curves for testing the neural network. Simulations indicated that the model had a high accuracy of predictions. The model ran simply, and the training and test data can be acquired easily. This study provides a new method for developing the cyclic constitutive model of super-elastic SMA. |
Key words: | disaster prevention and reduction engineering; shape memory alloy; super-elasticity; back-propagation neural network; constitutive model |
发表期数: | 2011年7月第14期 |
引用格式: | 任文杰,贾茹,韩雷. 超弹性形状记忆合金循环特性的BP神经网络模型研究[J]. 中国科技论文在线精品论文,2011,4(14):1283-1288. |

请您登录
暂无评论