论文标题
部分可观测时空混沌系统的无模型预测
DUG Insight: A software package for big-data analysis and visualisation, and its demonstration for passive radar space situational awareness using radio telescopes
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。
As the demand for software to support the processing and analysis of massive radio astronomy datasets increases in the era of the SKA, we demonstrate the interactive workflow building, data mining, processing, and visualisation capabilities of DUG Insight. We test the performance and flexibility of DUG Insight by processing almost 68,000 full sky radio images produced from the Engineering Development Array (EDA2) over the course of a three day period. The goal of the processing was to passively detect and identify known Resident Space Objects (RSOs: satellites and debris in orbit) and investigate how radio interferometry could be used to passively monitor aircraft traffic. These signals are observable due to both terrestrial FM radio signals reflected back to Earth and out-of-band transmission from RSOs. This surveillance of the low Earth orbit and airspace environment is useful as a contribution to space situational awareness and aircraft tracking technology. From the observations, we made 40 detections of 19 unique RSOs within a range of 1,500 km from the EDA2. This is a significant improvement on a previously published study of the same dataset and showcases the flexible features of DUG Insight that allow the processing of complex datasets at scale. Future enhancements of our DUG Insight workflow will aim to realise real-time acquisition, detect unknown RSOs, and continue to process data from SKA-relevant facilities.
