采用SVM的磨粒分类识别方法研究

为克服传统磨粒识别分类器训练时需要大量特征样本的缺点,设计一种基于多元支持向量机(Multi-Support Vector Machine,Multi-SVM)的磨粒识别分类器.支持向量机(SVM)是一种新的机器学习方法,在小样本和高维二元分类方面有非常突出的优点.实验证明,依据此优点设计的多元支持向量机磨粒分类器模型,不仅可以在小样本情形下对模型进行快速训练,而且可以快速识别多种磨粒类型,同时识别率也比传统的神经网络方法有较大提高,从而达到了提高设备监测和故障诊断效率的目的.

Debris recognition method based on multi-support vector machine

For solving the defect of traditional classificatory with many samples,a new classificatory recognizing debris based on Multi-SVM(Multi-Support Vector Machine)is proposed.SVM is a new machine study method which has excellent advantages in small-sample and multi-dimension binary classification.The new Multi-SVM classificatory can be studied in few samples rapidly to recognize several kinds of new debris.At the same time,the experiments showed that recognizing correct rate increased more greatly compared with...