Machine learning applied to laser-plasma interactions.

机器学习应用于激光-等离子体相互作用。

• 批准号: 2782834
• 金额: --
• 依托单位: University of York
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2782834
• 关键词: Machine; learning; applied; laser; plasma

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.

随着高重复频率激光系统的出现，激光等离子体物理学进入了一个新的领域，可以收集大量的数据，并用于测试和完善我们对潜在的复杂非线性现象的理解。新的BISHOP包装器（由斯特拉斯克莱德大学的合作者开发）围绕EPOCH粒子细胞代码自动化运行大量模拟的过程，因此提供了一个理想的工具来探索参数空间，以与产生大量数据的实验进行比较。与Strathclyde合作，我们将使用这个框架来研究激光固体相互作用中的粒子加速和辐射产生，BISHOP在参数空间的有趣区域中进行磨练。学生将通过实验研究这些机制，重点是高重复率带来的诊断挑战。EPOCH中将包含综合诊断，以帮助我们设计实际诊断并了解其要求。例子包括电子和离子光谱仪和X射线诊断。