A Second Order Approximation for the Caputo Fractional Derivative
Abstract: When $0<\alpha<1$, the approximation for the Caputo derivative $$y{(\alpha)}(x) = \frac{1}{\Gamma(2-\alpha)h\alpha}\sum_{k=0}n \sigma_k{(\alpha)} y(x-kh)+O\bigl(h{2-\alpha}\bigr),$$ where $\sigma_0{(\alpha)} = 1, \sigma_n{(\alpha)} = (n-1){1-a}-n{1-a}$ and $$\sigma_k{(\alpha)} = (k-1){1-\alpha}-2k{1-a}+(k+1){1-\alpha},\quad (k=1...,n-1),$$ has accuracy $O\bigl(h{2-\alpha}\bigr)$. We use the expansion of $\sum_{k=0}n k\alpha$ to determine an approximation for the fractional integral of order $2-\alpha$ and the second order approximation for the Caputo derivative $$y{(\alpha)}(x) = \frac{1}{\Gamma(2-\alpha)h\alpha}\sum_{k=0}n \delta_k{(\alpha)} y(x-kh)+O\bigl(h{2}\bigr),$$ where $\delta_k{(\alpha)} = \sigma_k{(\alpha)}$ for $2\leq k\leq n$, $$\delta_0{(\alpha)} = \sigma_0{(\alpha)}-\zeta(\alpha-1), \delta_1{(\alpha)} = \sigma_1{(\alpha)}+2\zeta(\alpha-1),\delta_2{(\alpha)} = \sigma_2{(\alpha)}-\zeta(\alpha-1),$$ and $\zeta(s)$ is the Riemann zeta function. The numerical solutions of the fractional relaxation and subdiffusion equations are computed.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.