论文标题

部分可观测时空混沌系统的无模型预测

Real time reconstruction of the fast electron spectrum from high intensity laser plasma interaction using gamma counting technique

论文作者

Zavorotnyi, A., Savel'ev, A.

论文摘要

储层计算是预测湍流的有力工具,其简单的架构具有处理大型系统的计算效率。然而,其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间,然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用,该系统模拟了湍流的若干特征。我们表明,使用径向基函数作为非线性投影器,即使只有部分观测并且不知道控制方程,也能稳健地捕捉复杂的系统非线性。最后,我们表明,当测量稀疏、不完整且带有噪声,甚至控制方程变得不准确时,我们的网络仍然可以产生相当准确的预测,从而为实际湍流系统的无模型预测铺平了道路。

X-ray and gamma fluxes from the high intensity laser-plasma interaction are extremely short, well beyond temporal resolution of any detectors. If laser pulses come repetitively, the single photon counting technique allows to accumulate the photon spectra, however, its relation to the spectrum of the initial fast electron population in plasma is not straightforward. We present efficient and fast approach based on the Geant4 package that significantly reduces computer time needed to re-construct the high energy tail of electron spectrum from experimental data accounting for the pileup effect. Here, we first tabulate gamma spectrum from monoenergetic electron bunches of different energy for a given experimental setup, and then compose the simulated spectrum. To account for the pileups, we derive analytical formula to reverse the data. We also consider errors coming from the approximation of the initial electron spectrum by the sum of monoenergetic impacts, the finite range of the electron spectrum, etc. and give estimates on how to choose modelling parameters to minimize the approximation errors. Finally, we present an example of the experimental data treatment for the case of laser-solid interaction using 50 fs laser pulse with intensity above 1018 W/cm2.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源