铁道科学与工程学报2018，Vol.15Issue(2)：328-335,8.
基于响应面模型的钢轨打磨廓形预测方法
A prediction method of rail profile based on response surface model
摘要
Abstract
Rail grinding is significant for restoring the rail profile. Due to the difference of initial rail profiles and grinding conditions, it is difficult to accurately predict the grinding profile of the rail formed by the combination of multiple grinding wheels before grinding in current approaches. Therefore, a new prediction method for grinding rail profile was proposed to solve this problem based on response surface model (RSM). After the collection of discrete points data of rail profiles, the method of cubic spline interpolation was brought up to illustrate the profile of rail before grinding. Taking the grinding power and the inclination angle of the grinding wheel as design variables, then, a third-order response surface model was established with the amount of grinding being the response. Based on the formation mechanism of the rail grinding profile, the numerical calculation scheme of the grinding profile was designed, and the grinding profile of the rail under the combination of multiple grinding wheels was worked out. According to some engineering tests, the accuracy and reliability of the method were verified according to the current rail grinding acceptance criteria.关键词
钢轨打磨量/三次样条插值/响应面模型/廓形预测Key words
the amount of rail grinding/cubic spline interpolation/response surface model/rail profile prediction分类
交通工程引用本文复制引用
刘冲,杨岳,丘文生,周鹏刚,易兵..基于响应面模型的钢轨打磨廓形预测方法[J].铁道科学与工程学报,2018,15(2):328-335,8.基金项目
国家自然科学基金资助项目(51605495) （51605495）
湖南省自然科学基金资助项目(2015JJ2168) （2015JJ2168）
