Direct-Drive Implosion Performance Optimization Using Gaussian Process Modeling and Reinforcement Learning

Finding a laser pulse shape that optimizes the Lawson parameter [1,2,3] (related to yield and ρR) for a given target is a challenging problem in inertial confinement fusion due to the predictive gap between simulations and experiments. Optimizing the yield for OMEGA experiments using a data-driven predictive machine-learning (ML) approach [4,5] has met with considerable success, but increasing the ρR has proven more challenging. It is likely that this is in part due to hydrodynamic instabilities but can also be caused by the increased sensitivity of the ρR to fine details of the shock timing and entropy spatial profile of the implosion, which in turn are highly sensitive to the front end of the laser pulse. If simulations used for implosion design [6] fail to capture the instability growth, shock transit, or adiabat-setting behavior of the implosion correctly, the response surface between simulations and experiments will sharply differ, making implosion optimization challenging with limited experimental data. We use Gaussian Processes [7] and Deep Q-learning [8] within an optimization framework that is developed on varying fidelities of simulation databases to find the optimum of a target design in a sample efficient manner during the course of a shot day. We present the use of this ML algorithm for optimizing experiments in synthetic space and on OMEGA experimental campaigns. This material is based upon work supported by the Department of Energy National Nuclear Security Administration under Award Number DE-NA0003856, the University of Rochester, and the New York State Energy Research and Development Authority.

 

[1] P. Y. Chang et al., Phys. Rev. Lett. 104, 135002 (2010).  

[2] B. K. Spears et al., Phys. Plasmas 19, 056316 (2012).  

[3] R. Betti, presented at the 24th IAEA Fusion Energy Conference, San Diego, CA, 8–13 October 2012.  

[4] V. Gopalaswamy et al., Phys. Plasmas 28, 122705 (2021).  

[5] A. Lees et al., Phys. Rev. Lett. 127, 105001 (2021).  

[6] J. Delettrez et al., Phys. Rev. A 36, 3926 (1987). 

[7] Gramacy, Robert B. Surrogates: Gaussian process modeling, design, and optimization for the applied sciences. CRC press, 2020 

[8] Mnih, Volodymyr, et al. "Playing atari with deep reinforcement learning." arXiv preprint arXiv:1312.5602 (2013).