Improved Memory Truncation Scheme for Quasi-Adiabatic Propagator Path Integral via Influence Functional Renormalization

TL;DR

This paper introduces a new memory truncation scheme for the iQuAPI method that selectively retains influential parts of the influence functional, enhancing accuracy and convergence for simulating non-Markovian quantum dynamics with long memory times.

Contribution

The paper proposes a novel memory truncation scheme using the density matrix renormalization group to improve iQuAPI's accuracy for long-memory quantum problems.

Findings

Faster convergence in simulations.

Enhanced accuracy over conventional truncation.

Effective for long-memory quantum dynamics.

Abstract

Accurately simulating non-Markovian quantum dynamics in system-bath coupled problems remains challenging. In this work, we present a novel memory truncation scheme for the iterative Quasi-Adiabatic Propagator Path Integral (iQuAPI) method to improve accuracy. Conventional memory truncation in iQuAPI discards all influence functional beyond a certain time interval, which is not effective for problems with a long memory time. Our proposed scheme selectively retains the most significant parts of the influence functional using the density matrix renormalization group algorithm. We validate the effectiveness of our scheme through simulations of the spin-boson model across various parameter sets, demonstrating faster convergence and improved accuracy compared to the conventional scheme. Our findings suggest that the new memory truncation scheme significantly advances the capabilities of iQuAPI for problems with a long memory time.