摘要
回转支承已在工程机械和风力发电等方面得到广泛应用。为了对其健康状态作出正确判断,采用经粒子群算法优化的支持向量机模型来对其寿命状态做出准确识别。寿命状态识别的关键问题是特征向量的提取。为了得到有效而又全面的寿命状态信息,从时域和时频域方面提取多个特征向量进行综合分析,从而实现了小样本数据下信息的最大挖掘。最后以回转史承全寿命实验对该方法进行检验,结果表明,该模型的效果优于传统的支持向量机以及单变量模型,具有实际工程应用价值。
Slewing bearing had been widely used in engineering machinery and wind power. In order to make the right judgments on their health status,the support vector machine (SVM) optimized by particle swarm algorithm model was proposed to make an accurate identification of the life state.The key problems of life state recognition was the feature vector extraction. To obtain an effective and comprehensive life state information of slewing bearing, multiple-feature vectors from time domain and time frequency domain were extracted, thus the information on small sample could be extracted as much as possible. Finally, the slewing bearing life experiments were used to test the model. Results demonstrated that the proposed model was better than the traditional SVM and univariate model, so it could be applied in the practical engineering.
出处
《南京工业大学学报(自然科学版)》
CAS
北大核心
2016年第1期56-61,67,共7页
Journal of Nanjing Tech University(Natural Science Edition)
基金
国家自然科学基金(51375222)
2014年度江苏省教育厅"青蓝工程"
关键词
回转支承
支持向量机
粒子群
寿命状态识别
slewing bearing
support vector machine
particle swarm optimization
life state recognition