Generalized variational calculus in terms of multi-parameters involving Atangana-Baleanu's derivatives and application (Q2180312)

The subject is the treatment of variational problems containing fractional derivatives. The definition of fractional derivative is \[ {}_aD^\alpha_t x(t) = \frac{B(\alpha)}{1 - \alpha} \int_a^t \frac{d}{ds} x(s) E_\alpha ( -\gamma(t - s)^\alpha)\, ds \] with $\gamma = \alpha/(1 - \alpha)$, $B(\alpha) = (1 - \alpha) + \alpha/ \Gamma(\alpha)$ for $\alpha > 0$, $B(0) = 1$ and $E_\alpha(z) = E_{\alpha, 1}(z)$, where $E_{\alpha, \beta}(z)$ is the Mittag-Leffler function \[ E_{\alpha,\beta}(z) = \sum_{n=0}^\infty \frac{z^n}{\Gamma(\alpha n + \beta)} \, . \] According to this definition ${}_a D^0_t x(t) = x(t) - x(a)$, ${}_a D^1_t x(t) = x'(t)$ as expected. There is a parallel definition where the differentiation sign is outside the integral. These are the left derivatives; for the right derivatives ${}_t D_b^\alpha$, the limits of integration are $t, b$. The author also defines fractional integrals and proves various results about the interplay of these with fractional derivatives. In particular, combining both concepts, another class of fractional derivatives ${}_a D_t^{\alpha. \beta} x(t)$, ${}_t D_b^{\alpha. \beta} x(t)$ is introduced. The main theorem is an Euler-Lagrange equation for the extremals of the functional \[ J(x) = \int_a^b L(t, x(t), {}_a D_t^{\alpha, \beta}x(t))\,dt \] with boundary conditions $x(a) = x_a$, $x(b) = x_b$ and a generalization for several variables. There are also results for the control problem of minimizing a performance index \[ J(u) = \int_0^1 F(t, x(t),u(t))\,dt \] with fractional differential dynamics \[ {}_0 D_t^{\alpha. \beta} x(t) = G(t, x(t), u(t)) \, , \quad x(0) = 0 \, . \] These are applied to a linear-quadratic problem where $F(t, x(t), u(t)) = q(t) x(t)^2 + r(t) u(t)^2$, $G(t, x(t), u(t)) = a(t)x(t) + b(t)u(t)$.