講演情報
[23p-A302-7]Orbital Optimize Method in ADAPT VQE
〇(M2)陽 建飛1、佐藤 健1、石川 顕一1、織茂 悠貴1 (1.東京大学大学院工学系研究科)
キーワード：
量子コンピュータ,変分量子固有値ソルバー,分子軌道の最適化
Adaptive Variational Quantum Eigensolver (ADAPT VQE)[1] is a quantum algorithm for simulating the electronic structure of molecular systems.The ADAPT VQE algorithm combines a Variational Quantum Solver (VQE) with an adaptive theorem optimization strategy to efficiently calculate the ground state energy of molecular systems. However, in the face of simulations of strongly correlated systems, ADAPT VQE often requires setting more stringent iterative termination conditions in order to be able to achieve chemical accuracy, which often entails a rapid increase in the cost of quantum hardware resources, which is still unrealistic for current NISQ devices.
To address challenge, this study uses an orbital optimization approach to improve the performance of ADAPT VQE.
Since the orbitals used in molecular simulations are often not optimal, optimization of the orbitals allows us to further improve the accuracy of the molecular ground state energy[2-3]. Under ansatz given by ADAPT VQE, the basis of the Hamiltonian is classically optimized like a variational parameter to find the optimal orbitals. It is important to emphasize that the orbital optimization method transforms the orbital integrals by a classical computer without introducing new quantum gates.This is certainly more friendly for NISQ devices.
The orbital optimize method can be used to further improve the accuracy of a given ansatz, and it is certainly a more economical way than generating a deeper ansatz.
To address challenge, this study uses an orbital optimization approach to improve the performance of ADAPT VQE.
Since the orbitals used in molecular simulations are often not optimal, optimization of the orbitals allows us to further improve the accuracy of the molecular ground state energy[2-3]. Under ansatz given by ADAPT VQE, the basis of the Hamiltonian is classically optimized like a variational parameter to find the optimal orbitals. It is important to emphasize that the orbital optimization method transforms the orbital integrals by a classical computer without introducing new quantum gates.This is certainly more friendly for NISQ devices.
The orbital optimize method can be used to further improve the accuracy of a given ansatz, and it is certainly a more economical way than generating a deeper ansatz.