改进Kriging方法在螺旋桨测量中的应用  

Improvement for Kriging Method in Applications of Blade Measurement

作　　者：李斌[1] 李沨[1] 陈福盈 付志凤 

机构地区：[1]华中科技大学国家数控系统工程技术研究中心,武汉430074 [2]镇江中船瓦锡兰螺旋桨有限公司,镇江212009

出　　处：《中国机械工程》2013年第13期1709-1714,共6页China Mechanical Engineering

基　　金：国家高技术研究发展计划(863计划)资助项目(2002AA424012);国家重点基础研究发展计划(973计划)资助项目(2011CB706803)

摘　　要：螺旋桨叶片在测量过程中因桨叶重叠而产生测量盲区,需要对盲区点坐标进行估计,估计精度直接影响桨叶的后期加工。借助Kriging插值算法的基本思想,提出一种基于混合数据(测量数据与设计数据)的改进Kriging插值算法。插值模型中,充分考虑螺旋桨截面线造型特点,使用样条函数构建回归模型;以测量数据构建模型主体,利用设计数据给出测量盲区曲线延展趋势。桨叶重叠区域的实验结果表明,该方法运行高效,估计精度满足企业使用要求。Some unknown points should be estimated to solve blind measurement area aroused by blade overlap in the process of propeller blade measurement, and the estimation results had an impor- tant effect on the later processing of blades. By means of the basic idea of Kriging interpolation algo- rithm, a improved Kriging interpolation algorithm based on mixture data （measurement data and de- sign data） was put forward. This new model employed spline method as the regression function, and whose subject was made up of measurement data. Besides, the characteristics of propeller section curve were considered adequately and the extension trend of the curve in the blind measurement area would be come up with design data. The experimental results show that the approach is of operation efficien- cy and the estimation results will satisfy application requirements of the enterprises.

关 键 词：控制点 KRIGING插值 样条曲线 回归函数 相关系数 

分 类 号：TH161[机械工程—机械制造及自动化] TB92[一般工业技术—计量学]