摘要
为了去除轮对压装压力信号中的异常点,获得真实的压装曲线,提出了一种异常点数据处理算法,即增量均值法,该算法结合压装压力信号的特点,能较好地识别并剔除异常点数据。在轮对压装过程中的实际应用表明,该算法的效果较好,可以消除干扰对轮对压装曲线的影响,从而减少对压装质量的误判。
In order to remove outliers from press signals acquired in press fitting process and obtain true press fitting curve, a new algorithm,incremental mean algorithm was presented. According to the traits of press signals,the algorithm can distinguish and remove outliers effectively. During practical application process ,the algorithm has received very good effects, eliminated misjudgments of press fitting quality.
出处
《武汉理工大学学报（信息与管理工程版）》
CAS
2013年第2期211-214,共4页
Journal of Wuhan University of Technology：Information & Management Engineering
基金
湖北省科学技术研究与开发基金资助项目(201097289)
关键词
轮对压装曲线
异常点
增量均值法
press fitting curve
outliers
incremental mean algorithm
