基于小波神经网络的滚动轴承故障诊断方法研究

摘要

This paper using wavelet analysis technology and neural network technology combined to realizing accuratefault feature extraction ,which through the analysis wavelet technology of vibration signal denoising, the fault information- signal feature representation of bearing frequency correlation analysis. Using the neural network which has the the ability of self-learning to improve the robustness and reliability of bearing fault diagnosis, in the statistical properties of the background noise is unknown, so that construction a low cost, high reliability of therolling bearing fault diagnosis and analysis system.The experimental results show that, the detection system in different training signal position ,signal acquisition position, bearing damage degree, training the extent of the damage detection of bearing under different circumstances, have good ability of detection and recognition,indicating that the system has good robustness, this method is effective and feasible.%将小波分析技术与神经网络技术相结合，通过小波分析技术对振动信号进行滤波消噪，构建时频关联分析的信号特征表征轴承的故障信息，实现准确的故障特征提取。运用神经网络方法具有的网络自适应能力，自学习能力，在背景噪声统计特性未知的情况下，提高轴承故障诊断系统的鲁棒性和可靠性，构建低成本、高可靠的滚动轴承故障分析诊断系统，实验结果表明，系统在训练信号、检测信号采集位置不同，训练轴承受损程度、检测轴承受损程度不同的情况下，均具有良好的检测识别能力，说明系统具有良好的鲁棒性，该方法有效可行。