• CN:11-2187/TH
  • ISSN:0577-6686

›› 2009, Vol. 45 ›› Issue (11): 106-112.

• 论文 • 上一篇    下一篇

扫码分享

基于改进Bussgang算法的多机组故障诊断仿真和试验

周勃;陈长征;费朝阳   

  1. 沈阳工业大学振动噪声研究所
  • 发布日期:2009-11-15

Simulation and Experiment of Multi-units Fault Diagnosis Based on Improved Bussgang Algorithm

ZHOU Bo;CHEN Changzheng;FEI Chaoyang   

  1. Institute of Vibration and Noise, Shenyang University of Technology
  • Published:2009-11-15

摘要: 针对非平稳信号改进Bussgang算法,借助参考噪声抵消包含在混合信号中的噪声,最小化输出信息的广义能量来获得最好的消噪效果;通过计算机仿真确定适合的非线性函数,仿真结果发现随着延迟系数的增大,收敛误差函数值减小,收敛性能增强;试验模拟了两个故障和三个故障的情况,分离结果与基于ICA、Bussgang算法的结果比较,验证改进算法对多机组含噪非平稳信号分离的有效性。

关键词: 多故障, 盲解卷, 振动诊断

Abstract: The Bussgang algorithm is improved for the non-stationary signals with the aid of the reference noise to eliminate the noise in the composite signals and to minimite the generalized energy of output information in order to obtain the best denoising effect. Through the computer simulation, the suitable nonlinear function is determined. It is found from the simulation results that the function value of convergence error decreases the convergence performance is enhanced along with delay factors enlargement. The situations of two-faults and three-faults are tested, and the separation results are compared with the results based on ICA and the Bussgang algorithm, which indicates the improved algorithm is suitable to separate the non-stationary signals of the multi-units.

Key words: Blind deconvolution, Multi-faults, Vibration diagnosis

中图分类号: