• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Mechanical Engineering and Mechanics vol.1; 20051026-28; Nanjing(CN) >Recognition of Pump State by RBF Neural Network
【24h】

Recognition of Pump State by RBF Neural Network

机译：通过RBF神经网络识别泵状态

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

On the basis of the traditional BP and RBF neural network, a new algorithm―user-defined step radial basis function was developed to monitor and recognize the three state (normal, wear and abnormal state) of vacuum air press pump ZYB03-60, then diagnosed the fault of it. The feature was extracted by sorting, learning and discriminating of the vibration signal. The frequency factor was taken into account in the new method. Therefore, the entire initial central node was adjusted and moved to the data area. This avoided the "dead center" which came from the traditional method with the principle of the nearest distance. In addition, the algorithm punished the competitor of the winning node and realized the choice of node automatically. Results indicate that the three states are presented in different area. The node reaches steady state with training 400 times and this is much faster than the traditional (1000 times of traditional).
机译：在传统的BP和RBF神经网络的基础上，开发了一种新的算法-用户定义的阶跃径向基函数来监视和识别真空空气压力泵ZYB03-60的三种状态（正常，磨损和异常状态），然后诊断出它的故障。通过对振动信号进行分类，学习和区分来提取特征。在新方法中考虑了频率因子。因此，整个初始中心节点均已调整并移至数据区域。这样就避免了以“最近距离”为原则的传统方法产生的“死点”。另外，该算法惩罚了获胜节点的竞争对手，并自动实现了节点的选择。结果表明这三种状态分别出现在不同的区域。节点经过训练400次达到稳定状态，这比传统节点（传统节点的1000倍）快得多。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号