线路电磁场测量数据的移动最小二乘拟合处理

Electromagnetic Measure Data Processing by Using Moving Least Square Method in Power Transmission Lines

为可靠评估输电线路周围电磁环境,针对有限的实测数据,将移动最小二乘近似方法引入到测量数据的拟合处理中。阐述了移动最小二乘逼近的原理,剖析其对数据移动加窗处理本质,给出了MLS应用于拟合的实施方案。具体以线路磁场测量数据为例,分别进行了分段拟合和MLS拟合处理,通过拟合曲线和误差的对比分析,反映MLS拟合具有参数调整方便、拟合曲线的连续性和光滑性好以及误差小的特点和优势。同时对权函数进行研究,给出参数选取的原则和方法。

It is difficult to evaluate the electromagnetic environment around power transmission lines credibly by using limited measurement data,therfore,the moving least square method in fitting processing is proposed which is aimed to solve such problem.After a detailed discussion of moving least square approximation＇s principle,the essential of MLS is analyzed.The MLS is in fact a kind of windowed processing on measurement data.And the complete implementation scheme of MLS is given subsequently.Then the measurement data of power transmission lines are numerically processed by subsection least square method and moving least square method separately.The comparison between the two methods consists of fitting curves＇ characteristics and fitting errors.Numerical results show that the MLS method has the properties of flexible adjustability on parameters,and the curve fitting is of good continuity and smoothness while with less fitting errors.The selection method of MLS＇ parameters are also studied,and some selection principles and methods are provided.The results show that the moving least square method not only is feasible in data post-processing,but also takes great advantages in such areas.