PolyFormer: learning efficient reformulations for scalable optimization under complex physical constraints
PolyFormer: 为在复杂物理约束下可扩展优化学习高效重述
机构 * School of Electrical and Electronic Engineering(电气电子工程学院) ; North China Electric Power University(华北电力大学) ; Department of Electrical Engineering(电气工程系) ; Tsinghua University(清华大学) ; School of Automation(自动化学院) ; Beijing Institute of Technology(北京理工大学) ; Department of Industrial Engineering(工业工程系) ; Laboratory for Systems and Control(系统与控制实验室) ; École Polytechnique Fédérale de Lausanne (EPFL)(瑞士联邦理工学院(EPFL))
AI总结 PolyFormer通过将物理和几何知识转化为高效多边形重述,实现了在复杂约束下的高效优化,显著提升计算效率和内存利用率。
Comments Code availability: All the data and code are made openly available at https://github.com/wenyl16/PolyFormer