Unleashed from Constrained Optimization: Quantum Computing for Quantum Chemistry Employing Generator Coordinate Method

arxiv(2023)

引用 0|浏览2
暂无评分
摘要
Hybrid quantum-classical approaches offer potential solutions for quantum chemistry problems, but they also introduce challenges such as the barren plateau and the exactness of the ansatze. These challenges often manifest as constrained optimization problems without a guarantee of identifying global minima. In this work, we highlight the interconnection between constrained optimization and generalized eigenvalue problems, using a unique class of non-orthogonal and overcomplete basis sets generated by Givens rotation-type canonical transformations on a reference state. Employing the generator coordinate approach, we represent the wave function in terms of these basis sets. The ensuing generalized eigenvalue problem yields rigorous lower bounds on energy, outperforming the conventional variational quantum eigensolver (VQE) that employs the same canonical transformations in its ansatze. Our approach effectively tackles the barren plateau issue and the heuristic nature of numerical minimizers in the standard VQE, making it ideal for intricate quantum chemical challenges. For real-world applications, we propose an adaptive scheme for selecting these transformations, emphasizing the linear expansion of the non-orthogonal basis sets. This ensures a harmonious balance between accuracy and efficiency in hybrid quantum-classical simulations. Our analysis and suggested methodology further broaden the applications of quantum computing in quantum chemistry. Notably, they pave the way for alternative strategies in excited state computation and Hamiltonian downfolding, laying the groundwork for sophisticated quantum simulations in chemistry.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要