AI 中文总结
本文提出基于傅里叶神经算子（FNO）的框架，开发FNO-SPMP方法，在水合氢离子分子的高维子空间中实现量子控制序列逆设计，显著提升成功率并缩短生成时间。
AI 中文摘要
量子最优控制是调控量子动力学的关键工具，但其计算成本会随希尔伯特空间维度的增加而快速增长。本文提出一种基于傅里叶神经算子（Fourier Neural Operator, FNO）的框架，用于学习高维分子量子动力学并加速控制协议的逆设计。给定初始分子布居分布、激光频率和偏振，FNO预测分子运动布居动力学的速度比使用CUDA-Q Dynamics的GPU加速数值传播快达10^7倍。利用这一快速且可微的代理模型，我们开发了FNO随机脉冲测量规划器（FNO-SPMP），该规划器构建脉冲序列以纯化初始的混合玻尔兹曼分布。我们在20K下的水合氢离子分子的888维子空间中演示了该协议，实现了0.98的目标态布居，序列成功率高达86.2%。在共享离散控制空间中，FNO-SPMP的成功率约为强化学习基线的两倍，同时使用的量子控制脉冲数量约为其一半，并将脉冲序列生成时间从约10小时缩短至10-20分钟。这些结果表明，算子学习代理模型可在希尔伯特空间过大而无法进行常规直接优化的量子系统中实现逆设计。
英文摘要
Quantum optimal control is a key tool for steering quantum dynamics, but its computational cost grows rapidly with the Hilbert space dimension. Here, we introduce a Fourier Neural Operator (FNO)-based framework for learning high dimensional molecular quantum dynamics and accelerating the inverse design of control protocols. Given an initial molecular population distribution, laser frequency, and polarization, the FNO predicts molecular-motional population dynamics up to $10^7$ times faster than GPU-accelerated numerical propagation with CUDA-Q Dynamics. Using this fast and differentiable surrogate, we develop the FNO stochastic pulse-measurement planner (FNO-SPMP), which constructs pulse sequences to purify an initially mixed Boltzmann distribution. We demonstrate the protocol in an 888-dimensional subspace of the hydronium molecule at 20 K, achieving a target-state population of 0.98 with a sequence success rate of up to 86.2%. In a shared discrete control space, FNO-SPMP achieves nearly twice the success rate of a reinforcement-learning baseline while using roughly half as many quantum control pulses and reducing pulse-sequence generation time from approximately 10 hours to 10-20 minutes. These results show that operator-learning surrogates can enable inverse design in quantum systems whose Hilbert spaces are too large for conventional direct optimization.
Comments19 Pages, 6 Figures
