采用非线性梯度共轭法的粗糙表面数字化模拟  

Numerical simulation of rough surfaces with non-linear conjugate gradient method

作　　者：炊明伟[1] 冯有前[2] 李培林[1] 徐晓东[1] 马军林[1] 

机构地区：[1]空军工程大学防空反导学院,西安710051 [2]空军工程大学理学院,西安710051

出　　处：《吉林大学学报（工学版）》2014年第1期241-245,共5页Journal of Jilin University:Engineering and Technology Edition

基　　金：国家自然科学基金项目(51075395)

摘　　要：为满足表面工程、结合面特性等研究领域对粗糙表面数据的大量需求,提出了一种新的三维非高斯粗糙表面数字化模拟方法。该方法根据设定的粗糙表面自相关函数,构造出求解转换矩阵的非线性方程组,并采用改进的非线性共轭梯度法进行求解;根据设定的峰度、偏斜度等粗糙表面高度参数,结合Johnson转换系统生成三维非高斯粗糙表面形貌。仿真结果表明,本文方法模拟的粗糙表面偏斜度、峰度误差小于3%,自相关函数模拟值与理论值拟合效果较好,能够根据设定的高度参数和自相关函数实现粗糙表面较准确的模拟。To meet the demands of large data of rough surfaces in the research of surface engineering and the characteristics of joint surfaces, a new numerical simulation method of 3D non-gaussian rough surfaces is proposed. Based on the assumed utocorrelation function （ACF）, the non linear equations are constructed for computing transformation matrix, and the improved non-linear conjugate gradient method is used to solve the equations. Then according to the assumed skewness, kurtosis and other statistical parameters of the rough surface, the 3D non-gaussian rough surfaces are generated by combining Johnson transformation system. The simulation results show that, the skewness and kurtosis relative errors of the simulated rough surfaces are less than 3G, the fitting result of ACF curves of simulated rough surfaces with the theoretical ACF is good. Using this method the rough surfaces with the given parameters of height and autocorrelation function can be accurately simulated.

关 键 词：信息处理技术 粗糙表面 数字化模拟 非线性梯度共轭法 自相关函数 

分 类 号：TN911[电子电信—通信与信息系统]