基于振动谱图像识别的智能故障诊断  

Intelligent Fault Diagnosis Using Image Recognition of Vibration Spectrogram

作　　者：林勇[1] 胡夏夏[1] 朱根兴[1] 钱少明[1] 

机构地区：[1]浙江工业大学之江学院,杭州310014

出　　处：《振动．测试与诊断》2010年第2期175-180,共6页Journal of Vibration,Measurement & Diagnosis

摘　　要：以滚动轴承为对象,提出了基于Hilbert包络分析和双谱分析的组合方法来提取振动信号的故障频率特征,进而生成双谱灰度图,利用双谱灰度图的灰度共生矩阵及其特征统计量来表征谱图特征。对该特征统计量进行主成分分析而得到的主分量,作为故障模式识别的输入向量。将用于故障模式分类的人工免疫网络分类算法,通过人工免疫网络对训练抗原进行学习形成记忆抗体网络,并计算检验抗原与记忆抗体的亲和力,按照正面选择的原理实现分类。在故障特征信号干扰严重的情况下,取得了较好的诊断准确率,验证了基于振动谱图识别的智能故障诊断方法的可行性。The Hilbert envelope analysis and bispectrum analysis were used to extract frequency components of vibration signals of a rolling bearing,and the gray level co-occurrence matrix(GLCM) and its characteristic statistics from bi-spectrum spectrogram were obtained.Then the characteristic statistics were further treated by the principal component analysis(PCA) to get principal components as the input vectors for fault pattern recognition.In an artificial immune network(AIN) system,the principal components were used as the antigens.A modified artificial immune network classification algorithm(AINCA) was introduced and used in the bearing fault diagnosis.Through the optimization of the fault antigens,the memory antibody nets formed and the fault classification was realized by the positive selection principle based on their affinity.The result indicates that the modified AINCA has a high accuracy of classification.

关 键 词：双谱分析 主成分分析 人工免疫网络分类算法 图像识别 智能故障诊断 滚动轴承 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TH133[自动化与计算机技术—控制科学与工程]