Becker Tobias, Netzer Ché, Eckardt André
Institut für Theoretische Physik, Technische Universität Berlin, Hardenbergstrasse 36, 10623 Berlin, Germany.
Phys Rev Lett. 2023 Oct 20;131(16):160401. doi: 10.1103/PhysRevLett.131.160401.
For the efficient simulation of open quantum systems, we often use quantum jump trajectories given by pure states that evolve stochastically to unravel the dynamics of the underlying master equation. In the Markovian regime, when the dynamics is described by a Gorini-Kossakowski-Sudarshan-Lindblad (GKSL) master equation, this procedure is known as Monte Carlo wave function approach. However, beyond ultraweak system-bath coupling, the dynamics of the system is not described by an equation of GKSL type, but rather by the Redfield equation, which can be brought into pseudo-Lindblad form. Here, negative dissipation strengths prohibit the conventional approach. To overcome this problem, we propose a pseudo-Lindblad quantum trajectory (PLQT) unraveling. It does not require an effective extension of the state space, like other approaches, except for the addition of a single classical bit. We test the PLQT for the eternal non-Markovian master equation for a single qubit and an interacting Fermi-Hubbard chain coupled to a thermal bath and discuss its computational effort compared to solving the full master equation.
为了有效地模拟开放量子系统，我们经常使用由纯态给出的量子跳跃轨迹，这些纯态随机演化以揭示基础主方程的动力学。在马尔可夫 regime 中，当动力学由戈里尼 - 科萨克夫斯基 - 苏达山 - 林德布拉德（GKSL）主方程描述时，此过程称为蒙特卡罗波函数方法。然而，在超弱系统 - 浴耦合之外，系统的动力学不是由 GKSL 型方程描述，而是由雷德菲尔德方程描述，该方程可转化为伪林德布拉德形式。在这里，负耗散强度禁止传统方法。为了克服这个问题，我们提出了一种伪林德布拉德量子轨迹（PLQT）分解。与其他方法不同，它不需要有效扩展态空间，除了添加一个经典比特。我们针对与热浴耦合的单个量子比特的永恒非马尔可夫主方程以及相互作用的费米 - 哈伯德链测试了 PLQT，并讨论了与求解完整主方程相比它的计算量。
