论文标题
部分可观测时空混沌系统的无模型预测
Monitoring pollution pathways in river water by predictive path modelling using untargeted GC-MS measurements
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。
A comprehensive approach to protect river water quality is needed within the European Water Framework Directive. Non-target screening of a complete chemical fingerprint of the aquatic ecosystem is essential, to identify chemicals of emerging concern and to reveal their suspicious dynamic patterns in river water. This requires a new combination of two measurement paradigms: the path of potential pollution should be traced through the river network, while there may be many compounds that make up this chemical composition - both known and unknown. Dedicated data processing of ongoing GC-MS measurements at 9 sites along the Rhine using PARAFAC2 for non-target screening, combined with spatiotemporal modelling of these sites within the river network using path modelling (Process PLS), provided a new integrated approach to track chemicals through the Rhine catchment, and tentatively identify known and as-yet unknown potential pollutants based on non-target screening and spatiotemporal behaviour.
