基于自适应邻域选取的局部投影降噪

提出一种以相点距离矩阵和信噪比为测度的自适应选取邻域参数的局部投影降噪方法.首先将一维时间序列重构到高维相空间,然后计算相空间中各个相点间的距离以构成相点距离矩阵,并据此设定参考相点的初始邻域半径和动态搜索步长,自适应地选择邻域大小,再利用局部几何投影方法消除噪声,迭代此过程,历史输出信噪比达到最大时获得优化邻域.

Local Projection Noise Reduction Method Based on Adaptive Neighborhood Selection

The local projection noise reduction method is presented based on adaptive neighborhood selection.The adaptive neighborhood selection was measured by a phase space distance matrix and RSN.First,one dimensional time series are embedded into a high dimensional phase space,the phase space distance matrix is calculated in order to describe the location between the phase points in high dimensional phase space,then the initial neighborhood size and dynamic step of each candidate phase point is set for adaptive neighborhood selection by a phase space distance matrix.Finally,local method is used to reduce noise.The optimal neighborhood size is obtained by iterative local projection when the out signal-to-noise ratio reaches the maximum.