Abstract
Variational quantum eigensolver (VQE), aiming at determining the ground state energy of a quantum system described by a Hamiltonian on noisy intermediate-scale quantum (NISQ) devices, is among the most significant applications of variational quantum algorithms (VQAs). However, the accuracy and trainability of the current VQE algorithm are significantly influenced due to the \emph{barren plateau} (BP), the non-negligible gate error and limited coherence time in NISQ devices. To tackle these issues, a gradient sensitive alternate framework with variable ansatz is proposed in this paper to enhance the performance of the VQE. We first propose a theoretical framework solving VA-VQE via alternately solving a gradient magnitudes related multi-objective optimization problem and the original VQE. Then, we propose a novel implementation based on the candidate tree based double $\epsilon$-greedy strategy and modified multi-objective genetic algorithm. As a result, the local optimum are avoided both in ansatz and parameter perspectives and the stability of output ansatz is enhanced. Furthermore, the experimental results show that, compared with the (arXiv:2010.10217) implementation, our framework is able to obtain the improvement of the error of the found solution, the quantum cost and the stability by up to 59.8%, 39.3% and 86.8%, respectively.
Abstract (translated)
URL
https://arxiv.org/abs/2205.03031
