Large deviations at various levels for run-and-tumble processes with space-dependent velocities and space-dependent switching rates

Abstract: One-dimensional run-and-tumble processes may converge towards some localized non-equilibrium steady state when the two velocities and/or the two switching rates are space-dependent. A long dynamical trajectory can be then analyzed via the large deviations at Level 2.5 for the joint probability of the empirical densities, of the empirical spatial currents and of the empirical switching flows. The Level 2 for the empirical densities alone can be then derived via the optimization of the Level 2.5 over the empirical flows. More generally, the large deviations of any time-additive observable can be also obtained via contraction from the Level 2.5, or equivalently via the deformed generator method and the corresponding Doob conditioned process. Finally, the large deviations for the empirical intervals between consecutive switching events can be obtained via the introduction of the alternate Markov chain that governs the series of all the switching events of a long trajectory.