Multi-timescale time encoding for CNN prediction of Fenna-Matthews-Olson energy-transfer dynamics
arXiv:2503.17430 · doi:10.1021/acs.jpclett.5c02761
Abstract
Machine learning simulations of open quantum dynamics often rely on recursive predictors that accumulate error. We develop a non-recursive convolutional neural networks (CNNs) that maps system parameters and a redundant time encoding directly to excitation-energy-transfer populations in the Fenna-Matthews-Olson complex. The encoding-modified logistic plus functions-normalizes time and resolves fast, transitional, and quasi-steady regimes, while physics-informed labels enforce population conservation and inter-site consistency. Trained only on reference trajectories generated with a Lindblad model in QuTiP, the network accurately predicts dynamics across a range of reorganization energies, bath rates, and temperatures. Beyond , the absolute relative error remains below 0.05, demonstrating stable long-time extrapolation. By avoiding step-by-step recursion, the method suppresses error accumulation and generalizes across timescales. These results show that redundant time encoding enables data-efficient inference of long-time quantum dissipative dynamics in realistic pigment-protein complexes, and may aid the data-driven design of light-harvesting materials.
8 pages, 5figures
References in corpus (6)
- Environment-Assisted Quantum Walks in Photosynthetic Energy Transfer
- Perspective: Numerically "exact" approach to open quantum dynamics: The hierarchical equations of motion (HEOM)
- A comparative study of different machine learning methods for dissipative quantum dynamics
- Machine learning for excitation energy transfer dynamics
- Influence of the coupled-dipoles on photosynthetic performance in a photosynthetic quantum heat engine
- Charge-transport enhanced by the quantum entanglement in the Photosystem II reaction center