Deep-Learning-Based Predictive Models for Laser Direct Drive at the Omega Laser Facility

The physics of inertial confinement fusion is rich and complex. Simulation codes that are used to design experiments are computationally expensive and lack the predictive capability required for extensive parameter exploration in search of a high-performing design for laser direct drive. In this work we use deep learning to build a fast emulator of experiments. To facilitate the development of the deep-learning model, an autoencoder is used to reduce the dimensionality of the input space. Two deep learning models are developed. One model is trained on a vast array of simulation data and is subsequently calibrated to expensive and limited experimental data using a technique known as “transfer learning.”[1] The other model is trained on a statistical model[2] and is subsequently calibrated using experimental data. A comparative study of the two predictive models is carried out. The models reproduce key experimental observables with high accuracy. Inference times on the DNN[JO1] models are unprecedented relative to the run time of simulation codes. The DNN models facilitate rapid exploration of a high dimensional input parameter space. Once high-performing designs are identified, high-fidelity simulations are used to understand the key physics of the design. This material is based upon work supported by the Department of Energy National Nuclear Security Administration under Award Number DE-NA0003856.

[1] K. D. Humbird et al., IEEE Trans. Plasma Sci. 48, 61 (2020).

[2] V. Gopalaswamy et al., Nature 565, 581 (2019).