1: \begin{abstract}
2: Many integrable stochastic particle systems
3: in one space dimension (such as TASEP --- Totally Asymmetric Simple Exclusion Process ---
4: and its various deformations, with a notable
5: exception of ASEP) remain integrable when
6: we equip each particle $x_i$ with its own jump rate parameter $\nu_i$.
7: It is a consequence of integrability that the distribution of each particle
8: $x_n(t)$
9: in a system started from the step initial configuration
10: depends on the parameters $\nu_j$, $j\le n$, in a symmetric way.
11: A transposition $\nu_n \leftrightarrow \nu_{n+1}$ of the parameters thus affects
12: only the distribution of $x_n(t)$.
13: For $q$-Hahn TASEP and its degenerations ($q$-TASEP and directed beta polymer)
14: we realize the transposition $\nu_n \leftrightarrow \nu_{n+1}$ as an explicit
15: Markov swap operator acting on the single particle $x_n(t)$.
16: For beta polymer, the swap operator can be interpreted as a simple
17: modification of the lattice on which the polymer is considered.
18: Our main tools are Markov duality and
19: contour integral formulas for joint moments.
20:
21: In particular, our constructions lead to a continuous time Markov
22: process $\mathsf{Q}^{(\mathsf{t})}$ preserving
23: the time $\mathsf{t}$
24: distribution of the $q$-TASEP
25: (with step initial configuration, where $\mathsf{t}\in \mathbb{R}_{>0}$ is fixed).
26: The dual system is a certain transient modification of the stochastic
27: $q$-Boson system. We identify asymptotic survival probabilities
28: of this transient process with
29: $q$-moments of the
30: $q$-TASEP, and
31: use this to show the convergence of the
32: process $\mathsf{Q}^{(\mathsf{t})}$
33: with arbitrary initial data
34: to its stationary distribution.
35:
36: Setting $q=0$, we recover the results about the usual TASEP established recently in
37: \cite{PetrovSaenz2019backTASEP} by a different approach
38: based on Gibbs ensembles of interlacing particles in two dimensions.
39: \end{abstract}
40: