Random-time processes governed by differential equations of fractional distributed order

Abstract

We analyze here different types of fractional differential equations, under the assumption that their fractional order $ν\in (0,1] $ is random\ with probability density $n(ν).$ We start by considering the fractional extension of the recursive equation governing the homogeneous Poisson process $N(t),t>0.$\ We prove that, for a particular (discrete) choice of $n(ν)$, it leads to a process with random time, defined as $N(% \widetilde{\mathcal{T}}_{ν_{1,}ν_{2}}(t)),t>0.$ The distribution of the random time argument $\widetilde{\mathcal{T}}_{ν_{1,}ν_{2}}(t)$ can be expressed, for any fixed $t$, in terms of convolutions of stable-laws. The new process $N(\widetilde{\mathcal{T}}_{ν_{1,}ν_{2}})$ is itself a renewal and can be shown to be a Cox process. Moreover we prove that the survival probability of $N(\widetilde{\mathcal{T}}_{ν_{1,}ν_{2}})$, as well as its probability generating function, are solution to the so-called fractional relaxation equation of distributed order (see \cite{Vib}%). In view of the previous results it is natural to consider diffusion-type fractional equations of distributed order. We present here an approach to their solutions in terms of composition of the Brownian motion $B(t),t>0$ with the random time $\widetilde{\mathcal{T}}_{ν_{1,}ν_{2}}$. We thus provide an alternative to the constructions presented in Mainardi and Pagnini \cite{mapagn} and in Chechkin et al. \cite{che1}, at least in the double-order case.