摘要
为了提高轴承状态监控的准确性,提出了一种基于模拟退火并可同时得到较好神经网络参数的新的优化方法。为验证所提方法的有效性,将实验台测得的滚动轴承振动信号作为研究样本,提取信号的特征。实验结果表明,该方法对轴承运行状态分类的准确率较高,可用于此类旋转机械的状态监控。
A simulated annealing approach is proposed in order to improve the classification accuracy,which can provide better parameter settings for network architecture of BPN.To verify the effectiveness of the proposed method,the roller bearing is tested under four operating conditions,five different shaft speeds and two load levels,and 52 features are extracted from the vibration signals of the tested bearing.The experimental results show that the proposed method can obtain a higher classification accuracy rate than other methods.Therefore,it is a promising approach to condition monitoring of rotating machinery.
出处
《机械工程与自动化》
2010年第3期120-121,126,共3页
Mechanical Engineering & Automation
关键词
状态监控
BP神经网络
模拟退火算法
轴承
condition monitoring
BP neural networks
simulated annealing algorithm
bearing
