摘要
以轨道车辆直流电机为研究对象,阐述了对电机多维数据进行特征选择的方法,并提出了一种将理想逼近法与余弦相似度和马氏距离相结合的电机性能评价方法。该方法与传统方法的对比结果表明,改进后的电机性能评价方法更具优越性和有效性。
Taking the traction motor of rail vehicle as the research object,the method of feature selection for multi-dimensional data of motor is described,and a motor performance evaluation method combining the ideal approximation method with cosine similarity and Markoy distance is propo sed.The comparison re sults of this method with the traditional method show that the improved motor performance evaluation method is more superior and effective.
作者
孔进
曹从咏
吴至锦
王晨
KONG Jin;CAO Congyong;WU Zhijing;WANG Cheng(Nanjing University,210094)
出处
《电机技术》
2022年第5期20-25,共6页
Electrical Machinery Technology
关键词
直流电机
性能评价
马氏距离
余弦相似度
DC motor
performance evaluation
Markov distance
cosine similarity
