Quantum neural network approach to Markovian dissipative dynamics of many-body open quantum systems
Cun Long,Long Cao,Liwei Ge,Qun-Xiang Li,YiJing Yan,Rui-Xue Xu,Yao Wang,Xiao Zheng
DOI: https://doi.org/10.1063/5.0220357
IF: 4.304
2024-08-22
The Journal of Chemical Physics
Abstract:Numerous variational methods have been proposed for solving quantum many-body systems, but they often face exponentially increasing computational complexity as the Hilbert space dimension grows. To address this, we introduce a novel approach using quantum neural networks to simulate the dissipative dynamics of many-body open quantum systems. This method combines neural-network quantum state representation with the time-dependent variational principle, both implemented via quantum algorithms. This results in accurate open quantum dynamics described by the Lindblad quantum master equation, exemplified by the spin-boson and transverse field Ising models. Our approach avoids the computational expense of classical algorithms and demonstrates the potential advantages of quantum computing for many-body simulations. To reduce measurement errors, we introduce a projection reset procedure, which could benefit other quantum simulations. In addition, our approach can be extended to simulate non-Markovian quantum dynamics.
chemistry, physical,physics, atomic, molecular & chemical