Machine learning applied to laser-plasma interactions.

点击翻译按钮获取中文摘要

With the advent of high repetition rate laser systems laser-plasma physics moves into a new regime,where large amounts of data can be collected and used to test and refine our understanding of theunderlying complex nonlinear phenomena. The new BISHOP wrapper (developed by collaborators atthe University of Strathclyde) around the EPOCH particle-in-cell code automates the process ofrunning a large number of simulations and so provides an ideal tool to explore parameter space forcomparison to experiments producing high volumes of data. In collaboration with Strathclyde, wewill use this framework to investigate particle acceleration and radiation generation in laser solidinteractions, with BISHOP honing in on interesting regions of parameter space. The student willinvestigate these regimes experimentally, focusing on the diagnostic challenge posed by highrepetition rate. Synthetic diagnostics will be included in EPOCH to help us design the realdiagnostics and understand their requirements. Examples include electron & ion spectrometers andx-ray diagnostics.