物理信息人工智能逆变器
系统与控制
2024-07-12 v4 系统与控制
摘要
本文设计了一种AI逆变器，它开创性地使用物理信息神经网络(PINN)来实现基于人工智能的构网型逆变器电磁暂态仿真(EMT)。贡献有三点：(1) 提出了一种基于PINN的AI逆变器；(2) 设计了一种增强的学习策略，即平衡自适应PINN；(3) 对AI逆变器的准确性和效率进行了广泛的验证和比较分析，以显示其相对于经典电磁暂态程序(EMTP)的优越性。
引用
@article{arxiv.2406.17661,
title = {Physics-Informed AI Inverter},
author = {Qing Shen and Yifan Zhou and Peng Zhang and Yacov A. Shamash and Roshan Sharma and Bo Chen},
journal= {arXiv preprint arXiv:2406.17661},
year = {2024}
}
备注
We are working on significantly expanding the research(methodology and test cases), and the current version does not accurately reflect our findings. Need more experiments to draw the conclusion. The experiments are still undergoing. We need more time to refine it. It is not ready to be public