论文标题
部分可观测时空混沌系统的无模型预测
The Son-Of-X-shooter (SOXS) Data-Reduction Pipeline
论文作者
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。
The Son-Of-XShooter (SOXS) is a single object spectrograph (UV-VIS & NIR) and acquisition camera scheduled to be mounted on the ESO 3.58-m New Technology Telescope at the La Silla Observatory. Although the underlying data reduction processes to convert raw detector data to fully-reduced science ready data are complex and multi-stepped, we have designed the SOXS Data Reduction pipeline with the core aims of providing end-users with a simple-to-use, well-documented command-line interface while also allowing the pipeline to be run in a fully automated state; streaming reduced data into the ESO Science Archive Facility without need for human intervention. To keep up with the stream of data coming from the instrument, there is the requirement to optimise the software to reduce each observation block of data well within the typical observation exposure time. The pipeline is written in Python 3 and has been built with an agile development philosophy that includes CI and adaptive planning.
