基于径向基函数的变量预测模型模式识别方法

发布时间：2018-05-21 10:39

本文选题：径向基函数(RBF) + 变量预测模式识别方法　；参考：《航空动力学报》2017年02期

【摘要】：针对变量预测模型模式识别方法中4种数学模型不足以反映特征值之间复杂关系的缺陷.因此,提出了一种基于径向基函数的变量预测模型(VPMRBF)模式识别方法,把提取的特征值输入到VPMRBF分类器中,然后通过训练样本建立反映特征值之间复杂关系的径向基函数预测模型,最后把测试样本的特征值作为径向基函数预测模型的输入,以预测误差平方和为依据完成分类.该方法充分有效地利用并且结合径向基函数和变量预测模式识别方法的优点,实现了故障特征提取到故障识别的全程诊断.滚动轴承故障诊断实验分析结果表明:与径向基神经网络、支持向量机和变量预测模式识别方法相比,VPMRBF的识别率分别提高了4.75%,1.75%和5.25%.
[Abstract]:The four mathematical models in the pattern recognition method of variable prediction model are not sufficient to reflect the complex relationship between the eigenvalues. Therefore, a variable prediction model based on radial basis function (RBF) is proposed, and the extracted eigenvalues are input into the VPMRBF classifier. Then the radial basis function prediction model reflecting the complex relationship between the eigenvalues is established through the training samples. Finally, the characteristic values of the test samples are taken as the input of the radial basis function prediction model, and the classification is completed on the basis of the sum of the square of prediction errors. This method makes full and effective use of the advantages of radial basis function (RBF) and variable prediction pattern recognition method, and realizes the whole process of fault diagnosis from fault feature extraction to fault identification. The experimental results of rolling bearing fault diagnosis show that compared with radial basis function neural network, support vector machine and variable prediction pattern recognition method, the recognition rate of VPMRBF is increased by 4.75% and 5.25%, respectively.
【作者单位】：湖南大学汽车车身先进设计制造国家重点实验室;安徽工业大学机械工程学院;
【基金】：国家重点研发计划(2016YFF0203400) 国家自然科学基金(51575168,51375152) 智能型新能源汽车国家2011协同创新中心资助项目湖南省绿色汽车2011协同创新中心资助项目
【分类号】：TP391.4

【参考文献】