一种求解误差数据近似导数的快速算法

A Fast Algorithm for Numerical Differentiation of Discrete Noisy Data

基于第一类Volterra积分方程,给出了一种在数据的误差水平未知的情况下,求已知数据的一阶近似导数的快速算法。数值试验表明该方法是有效的,数值稳定性高,计算量小。

Based on the first kind of Volterra integral equation, a fast algorithm for numerical differentiation is proposed, which can even works when the data noise level is unkown .The numerical test shows that this method possessed the advantage of high accuracy, good numerical stability and less computation.