摘要
针对以往模式识别方法的不足及特征数据存在"异常值"导致的模型失真问题,提出了基于特征评价的优化加权代理判别模型(Agent discriminate model based optimization weighted,ADMOW)模式识别方法。该方法根据同一状态类别中各特征值之间的对应关系(不同类别有不同的对应关系)建立数学预测模型,然后计算各特征值的类相似度评价指标,根据评价指标对特征值进行加权处理,从而削弱"异常值"对模型的影响,建立更加准确的代理判别模型,提高分类准确度。实验结果表明,经过特征加权处理的ADMOW方法对滚动轴承的状态识别具有更高的识别率。
In view of the shortcomings of the traditional pattern recognition methods and model distortion due to the outliers in the feature data,a pattern recognition method of agent discriminate model based optimization weighted(ADMOW)was proposed on the basis of characteristic evaluation.According to the corresponding relation between the characteristic values in the same category,the mathematical prediction model was established.Then,the similarity evaluation indexes of different eigenvalues were calculated.Meanwhile,the eigenvalues were processed with feature weighting according to the evaluation indexes so that the influence of the outliers was weakened.Finally,a more accurate agent classification model was established and the accuracy of classification was improved.The experiment results show that theADMOWmethod with feature weighting has a higher efficiency in rolling bearing status recognition.
作者
张建
潘海洋
郑近德
潘紫微
ZHANG Jian;PAN Haiyang;ZHENG Jinde;PAN Ziwei(School of Mechanical Engineering,Anhui University of Technology,Ma’anshan 243032,Anhui China)
出处
《噪声与振动控制》
CSCD
2018年第4期192-197,共6页
Noise and Vibration Control
基金
国家重点研发计划课题资助项目(2017YFC0805103)
国家自然科学基金资助项目(51505002)
安徽省自然科学基金资助项目
