摘要
模式识别的首要问题是特征提取,在特征提取过程中,是否可以找到有效区分不同模式的特征,对模式识别的准确率有至关重要的影响。特征向量在高维可视化过程中,每一类都有自己固有的几何形态,在原有特征的基础上,可以进一步从斜率的角度反映每一类的形状,为此提出一种基于斜率关联度的模式识别方法,利用UCI机器学习数据库中的Iris数据集进行了模式识别效果分析,得到了较高的识别准确率。将基于斜率关联度的模式识别方法应用于滚动轴承故障诊断,同样得到了较理想的故障诊断效果。
Pattern recognition is key to feature extraction and the effective extraction of features of different patterns is vitally important for the accuracy of pattern recognition. In the high-dimension visualization process, each class of feature vectors has its inherent geometric shape. Therefore, we propose our pattern recognition method that uses the slope relational degree. We use the Iris data set in the UCI machine learning database to analyze the pattern recognition effectiveness, thus obtaining a highly accurate pattern recognition rate. Finally, we apply our pattern recognition method to the fault diagnosis of a rolling bearing and also obtain ideal fault diagnosis results.
出处
《机械科学与技术》
CSCD
北大核心
2012年第9期1500-1503,共4页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(51175316)
教育部博士点基金项目(20103108110006)
上海市人才发展资金项目(047)资助
关键词
斜率关联度
模式识别
滚动轴承
故障诊断
slope relational degree
pattern recognition
rolling bearing
fault diagnosis
