Power system operators are increasingly deploying Grid Enhancing Technologies (GETs) to mitigate operational challenges such as line and transformer congestion, and voltage violations. These technologies, including Network Topology Optimization (NTO), Variable Impedance Devices (VIDs), and Dynamic Line Rating (DLR), enhance system flexibility and enable better utilization of existing network assets. However, as the deployment of multiple GETs grows, effective coordination among them becomes essential to fully realize their potential benefits. This paper presents a co-optimization framework that models and coordinates NTO, VID, and DLR within a unified optimization scheme to alleviate network congestion and minimize operational costs. The NTO formulation is developed using a node-breaker model, offering finer switching granularity and improved operational flexibility. The inclusion of VIDs introduces nonlinear and non-convex relationships in the optimization problem. DLR takes into account of weather conditions, primarily wind speed and ambient temperature, enabling adaptive utilization of transmission capacity. The proposed framework is validated on standard IEEE benchmark test systems, demonstrating its effectiveness under varying numbers and placements of impedance controllers.
A Perspective on the Algebra, Topology, and Logic of Electrical Networks
National and University Library in Zagreb(萨格勒布国家大学图书馆);School of Applied Mathematics and Informatics, University of Osijek(奥西耶克大学应用数学与信息学院);Faculty of Electrical Engineering and Computing, University of Zagreb(萨格勒布大学电气工程和计算学院)
Deep Learning Assisted End-to-End Synthesis of mm-Wave Passive Networks with 3D EM Structures: A Study on A Transformer-Based Matching Network
School of Industrial and Systems Engineering, Georgia Tech(佐治亚理工学院工业与系统工程系);School of Electrical and Computer Engineering, Georgia Tech(佐治亚理工学院电气与计算机工程系)
本文提出了一种深度学习辅助综合方法,用于直接端到端生成具有3D EM结构的RF/mm波无源匹配网络。与现有的方法,从目标电路元件值和目标拓扑结构合成EM结构不同,我们所提出的方法实现了直接合成的无源网络给定的网络拓扑结构,从期望的性能值作为输入。我们展示了建议的合成神经网络(NN)模型上的片上1:1变压器为基础的阻抗匹配网络。通过利用参数共享,合成NN模型成功地从输入阻抗和负载电容器中提取相关特征,并预测45 nm SOI工艺中的Transformer 3D EM几何形状,其将标准50 $Ω$负载与目标输入阻抗匹配,同时吸收两个负载电容器。作为概念验证,合成了几个示例Transformer几何形状,并在Ansys HFSS中进行了验证,以提供所需的输入阻抗。