欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出了一种基于局部特征尺度分解（Localcharacteristicscaledecomposition，LCD）和核最近邻凸包（Kernelnearestneighborconvexhull，KNNCH）分类算法的滚动轴承故障诊断方法。采用LCD方法对滚动轴承原始振动信号进行分解得到若干内禀尺度分量（Intrinsicscalecomponent，ISC），然后将这些ISC分量组成初始特征向量矩阵，再对该矩阵进行奇异值分解，提取奇异值作为故障特征向量并输入到KNNCH分类器，根据其输出结果来判断滚动轴承的工作状态和故障类型。LCD方法是一种新的自适应时频分析方法，非常适用于非平稳信号的处理，而KNNCH算法是一种基于核函数方法，并将凸包估计与最近邻分类思想相融合的模式识别算法，可直接应用于多类问题且需优化的参数只有核参数。实验分析结果表明，所提出的方法能有效地提取滚动轴承故障特征信息，而且在小样本的情况下仍能准确地对滚动轴承的工作状态和故障类型进行分类。同时，与支持向量机（Supportvec～tormachine，SVM）算法的对比分析结果表明，KNNCH算法的分类性能的稳定性要高于SVM算法。
英文摘要：
A rolling bearing fault diagnosis approach is proposed based on local characteristic-scale decomposition （LCD） and kernel nearest neighbor convex hull （KNNCH） classification algorithm. By using LCD, an original rolling bearing vibration signal could be adaptively decomposed into a number of intrinsic scale components （ISC）, and an initial feature vector matrix is automatically formed from these components. Then, by applying singular value decomposition technique to the initial feature vector matrix, singular values are obtained and regarded as the fault feature vector. Finally, KNNCH classifier accepts the fault feature vector as the input, and then the working condition and fault patterns of rolling bearing could be identified by the output of the classifier. LCD is a new adaptive time-frequency analysis method which very suits non-stationary signals process- ing. Additionally, KNNCH algorithm is a kernel-based pattern recognition approach which combines convex hull estimation and nearest neighbor classification rule. Contrast to support vector machine （SVM） algorithm, KNNCH algorithm could be di- rectly applied to multi-class tasks and the parameter needed to be optimized is only the kernel parameter. The analysis results from rolling bearing vibration signals show that the proposed approach can effectively extract the fault feature information and accurately classify the working conditions and fault patterns of rolling bearing even in the case of small samples. What＇s more, the comparative analysis results demonstrate that KNNCH algorithm gains more stable classification performance than SVM algorithm.
同期刊论文项目
同项目期刊论文
Rolling bearing fault diagnosis based on multi-scale fuzzy entropy and variable predictive model-bas
Normalized complex Teager energy operator demodulation method and its application to fault diagnosis
Velocity Consistency Analysis for Selecting Excellent Feature Points in Three-Dimensional Reconstruc
Solving Linear Coupled Fractional Differential Equations by Direct Operational Method and Some Appli
The Specular Exponent as a Criterion for Appearance Quality Assessment of Pearl-Like Objects by Arti
Parametric Shape Representation by a Deformable NURBS Model for Cardiac Functional Measurements[基于可变
Detection and amendment of shape distortions based on moment invariants for active shape models, [基于
Elman Fuzzy Adaptive Control for Obstacle Avoidance of Mobile Robots using Hybrid Force/Position Inc
Generalized empirical mode decomposition and its applications to rolling element bearing fault diagn
期刊信息
位置：