There has been tremendous recent interest in Direct Air Capture (DAC) systems. A key part of any DAC system are the multiple air intake units. In particular, the arrangement of such units for optimal capture and sequestration is critical. Accordingly, this work develops an easy to use model for a modular unit system, where an approximate flow field is computed for each unit and the aggregate flow field is developed by summing the fields from each unit. This allows for a modular framework that can be used for rapid simulation and design of an overall DAC system. The rapid rate at which these simulations can be completed enables the ability to explore inverse problems seeking to determine which parameter combinations can deliver the maximum sequestration of tracer plume particles for the minimum amount of energy input. In order to cast the objective mathematically, we set up an inverse as a Machine Learning Algorithm (MLA); specifically a Genetic MLA (G-MLA) variant, which is well-suited for nonconvex optimization. Numerical examples are provided to illustrate the framework.

中文翻译：

---

用于快速模拟模块化直接空气捕获系统的数字孪生


最近人们对直接空气捕获（DAC）系统产生了极大的兴趣。任何 DAC 系统的关键部分都是多个进气单元。特别是，此类单元的布置对于最佳捕获和封存至关重要。因此，这项工作为模块化单元系统开发了一种易于使用的模型，其中计算每个单元的近似流场，并通过对每个单元的场求和来开发聚合流场。这使得模块化框架可用于快速仿真和设计整个 DAC 系统。这些模拟的快速完成使得能够探索反问题，以确定哪些参数组合可以以最小的能量输入实现示踪羽流粒子的最大封存。为了以数学方式实现目标，我们建立了一个逆机器学习算法（MLA）；特别是遗传 MLA (G-MLA) 变体，它非常适合非凸优化。提供了数值示例来说明该框架。