An Artificial Intelligence Engineering System Analysis Assistant (Aiesaa) for auto-creation of integrated transmission-distribution grid models

分布式能源（DER）的整合，如太阳能光伏和电池系统，正在彻底改变电网，提供新的可能性和挑战。为了在考虑运行约束的同时准确地模拟聚合DER行为对电力传输的影响，有必要开发全面的综合输配电（TD）网络模型。然而，由于大量的系统连接到区域输电网，为每个配电系统创建全尺寸模型是不切实际的。在这个项目中，我们的主要目标是开发Aiesaa，一个专业智能（AI）助手，以改变创建紧凑和集成的TD网络模型的过程。我们的目标是克服劳动密集型的性质，可扩展性和模型转换的问题，以及当前的协同仿真方法所面临的通信挑战。Aiesaa将利用先进的机器学习技术来简化三个关键的建模任务。首先，它将有助于场景分类，允许人类专家专注于可以使用简化模型的非关键场景。其次，Aiesaa将采用元建模技术为关键场景选择和参数化降阶模型，在准确性和复杂性之间取得平衡。最后，Aiesaa将促进人在环模型集成，确保人工智能和专家之间的协作，以实现最佳的模型性能和复杂性。通过将人工智能的速度和准确性与人类专家的见解和经验相结合，Aiesaa将推出一种超越现有方法的工程模型创建新框架。这种方法可以自动执行常规任务和工作流程，使专家能够专注于需要他们专业知识的更高级别的活动。重要的是，人在回路方法确保人工智能作为合作者，而不是人类专业人员的替代品。Aiesaa的发展对计算效率和成本效益具有重要意义。通过降低模型复杂性和缩短开发时间，Aiesaa使紧凑的集成TD模型在独立的计算平台上使用成为可能。这减少了对昂贵基础设施的依赖，增强了数据安全性，并加快了模拟速度。此外，Aiesaa降低了建模人员的学习曲线，使他们能够专注于更高级别的任务，如工程系统设计和未来场景。项目完成后，我们计划与研究和工程界分享Aiesaa的原型，促进该领域的进步。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The integration of distributed energy resources (DERs), such as solar photovoltaic and battery systems, is revolutionizing power grids, offering new possibilities and challenges. To accurately model the impact of aggregated DER behaviors on power transmission while considering operational constraints, it is essential to develop comprehensive integrated transmission-distribution (T&D) network models. However, creating full-scale models for each distribution system is impractical due to the large number of systems connected to the regional transmission grid. In this project, our main objective is to develop Aiesaa, an Artificial-intelligence (AI) assistant, to transform the process of creating compact and integrated T&D network models. We aim to overcome the labor-intensive nature, scalability and model conversion issues, and communication challenges faced by current co-simulation approaches. Aiesaa will leverage advanced machine learning techniques to streamline three crucial modeling tasks. Firstly, it will assist in scenario classification, allowing human experts to focus on non-critical scenarios where simplified models can be used. Secondly, Aiesaa will employ meta-modeling techniques to select and parameterize reduced-order models for critical scenarios, striking a balance between accuracy and complexity. Lastly, Aiesaa will facilitate human-in-the-loop model integration, ensuring collaboration between AI and experts to achieve optimal model performance and complexity.By combining the speed and accuracy of AI with the insights and experiences of human experts, Aiesaa will introduce a novel framework for engineering model creation that surpasses existing methodologies. This approach automates routine tasks and workflows, freeing up experts to concentrate on higher-level activities that demand their expertise. Importantly, the human-in-the-loop approach ensures that AI serves as a collaborator rather than a replacement for human professionals. The development of Aiesaa has significant implications for computational efficiency and cost-effectiveness. By reducing model complexity and shortening development time, Aiesaa enables the use of compact integrated T&D models on standalone computing platforms. This reduces reliance on expensive infrastructure, enhances data security, and accelerates simulations. Additionally, Aiesaa reduces the learning curve for modelers, empowering them to focus on higher-level tasks such as engineering system design and future scenarios. Upon completion of the project, we plan to share a prototype of Aiesaa with the research and engineering community, fostering advancements in the field.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.