利用量子相干冷却逃离局部极小值
Escaping Local Minima with Quantum Coherent Cooling
AI总结:
针对经典优化易陷入局部极小值的问题,提出一种量子-经典混合算法,利用量子相干热浴和协同隧穿帮助系统穿越势垒,并在 cQED 中实现,数值显示其求解自旋玻璃具有量子优势。
摘要:
量子冷却已在量子计算中展现出潜力,它可以减少外部信号所需的控制通道数量。近期进展也支持在大规模系统中维持量子相干性的可能性。借助这一方案,经典算法困在代价函数局部极小值中的局限有望被克服。基于此,我们提出一种用于寻找全局极小值的量子-经典混合算法。当经典算法陷入停滞时,我们的方法利用量子相干冷却促进协同隧穿,以穿越能量势垒。经过编码的 Hamiltonian 系统表示代价函数,而处于基态的量子相干热浴充当散热器,吸收来自系统的能量。我们提出的方案可在使用量子腔的电路量子电动力学(cQED)系统中实现。所提供的数值证据证明了其在求解自旋玻璃问题时的量子优势。
英文摘要:
Quantum cooling has demonstrated its potential in quantum computing, which can reduce the number of control channels needed for external signals. Recent progress also supports the possibility of maintaining quantum coherence in large-scale systems. The limitations of classical algorithms trapped in local minima of cost functions could be overcome using this scheme. According to this, we propose a hybrid quantum-classical algorithm for finding the global minima. Our approach utilizes quantum coherent cooling to facilitate coordinative tunneling through energy barriers if the classical algorithm gets stuck. The encoded Hamiltonian system represents the cost function, and a quantum coherent bath in the ground state serves as a heat sink to absorb energy from the system. Our proposed scheme can be implemented in the circuit quantum electrodynamics (cQED) system using a quantum cavity. The provided numerical evidence demonstrates the quantum advantage in solving spin glass problems.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。