一类离散奇异褶积微分算子的优化方法

Optimal Method for a Kind of Discrete Singular Convolutional Differentiator

提出了一种优化正则化Shannon核离散奇异褶积微分算子的方法,使得优化后的正则化Shannon核离散奇异褶积微分算子在抑制Gibbs现象的同时,在光滑性、精度等方面也有较大的提高,并运用数值算例进行了试验对比,说明此优化方法的有效性,为褶积微分算子高效地运用到实际问题中提供了强有力的保证.

In this paper,a method for optimizing the discrete singular convolutional differentiator with regularized Shannon＇s kernel is proposed.The optimized discrete singular convolutional differentiator with regularized Shannon＇s kernel can suppress the Gibbs phenomenon and improve the smoothness and the accuracy.Comparisons of numerical experiments are made to show that the optimization method is effective.The appealing features of the optimized convolutional differentiator would make it more effective in dealing with practical issues.