Theory And Numerical Approximations Of Fractional Integrals And Derivatives -

$$ 0^CD^\alpha t f(t_n) \approx \frach^-\alpha\Gamma(2-\alpha) \sum_j=0^n-1 b_j \left[ f(t_n-j) - f(t_n-j-1) \right]$$

It is structured to move from foundational theory to computational methods, highlighting key challenges. 1. Introduction: Beyond Integer Order Classical calculus deals with derivatives and integrals of integer order. Fractional calculus (FC) generalizes these operations to arbitrary real (or complex) orders. While this generalization introduces powerful tools for modeling memory effects and non-local behavior in viscoelasticity, anomalous diffusion, signal processing, and control theory, it comes at a cost: fractional operators are inherently non-local . Consequently, numerical approximations are rarely straightforward extensions of their integer-order counterparts. 2. Foundational Theory: Definitions and Key Properties Unlike integer calculus, where the derivative is unique, several definitions of fractional derivatives exist. The choice depends on the problem's initial/boundary conditions and desired properties. 2.1 The Fractional Integral (Riemann–Liouville) The natural starting point is the Cauchy formula for repeated integration, generalized via the Gamma function $\Gamma(\cdot)$. For order $\alpha > 0$, the left-sided Riemann–Liouville fractional integral is: singular at $t=0$). For many problems

$$ aI^\alpha t f(t) = \frac1\Gamma(\alpha) \int_a^t (t-\tau)^\alpha-1 f(\tau) , d\tau$$ computational cost (dense vs. compressed history)

$$ a^GLD^\alpha t f(t_n) \approx h^-\alpha \sum_j=0^n \omega_j^(\alpha) f(t_n-j)$$ where the derivative is unique

The choice of numerical method in fractional calculus is a trade-off between physical fidelity (long memory), computational cost (dense vs. compressed history), and regularity of the solution (smooth vs. singular at $t=0$). For many problems, the short-memory principle or sum-of-exponentials acceleration is not a luxury—it is a necessity.