Robust fractional order differentiator

A robust fractional order differentiator is proposed to estimate the fractional order derivative of the given contaminated signal. Least square polynomial fitting technique is used to minimize the noise and Grünwald-Letnikov definition is used to calculate the fractional order derivative. A concise method is proposed to speed up the computation time and a parameter setting strategy is present for user. The frequency domain analysis demonstrates that the proposed differentiator essentially is a fractional order low-pass filter. The accuracy and robustness are evaluated by the experiments.