基于能量差的加权均值递归滤波算法

Research on weighted average recursive filter based on energy difference

对基于激光测量的管形零件内轮廓母线采样数据进行滤波.提出了基于能量差的加权均值递归滤波算法.该滤波方法中权值求取方法的思想是尽量给滤波窗口内能量频谱相对集中的采样数据赋予较大的权值,而给由于干扰引起的畸变数据赋予较小的权值.合适的滤波窗口长度通过仿真对比试验获得.这一思想的合理性通过仿真分析和实验测试得到验证,并表现出良好的效果.

In order to filter the data sampled from inner-contour generatrix of pipe figure measured by laser, a new effective weighted average recursive filter based on energy difference is proposed, In counting the weight of the filter proposed, the idea is that more weight will be distributed to those data in which the components of energy spectrum relatively concentrate and less weight to other aberration data disturbed by noise. Proper length of filter window is got from simulation results, The well perfor- mance of the filter verified by simulations and experiments indicates the idea＇s rationality.