New-generation optimization algorithms for engineering

新一代工程优化算法

• 批准号: 21K17710
• 负责人: Themelis Andreas
• 金额: $ 2万
• 依托单位: Kyushu University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Early-Career Scientists
• 财政年份: 2021
• 资助国家: 日本
• 起止时间: 2021-04-01 至 2024-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-21K17710/
• 关键词: Optimization algorithms; Nonconvex optimization; Adaptive algorithms; Autonomous driving; Decentralized control; Open-source software

The candidate established a new collaboration with the University of British Columbia (CA), including in the discussion a PhD student under his co-supervision that resulted in the preprint [1]. In addition, he has been invited for visiting periods at the Tokyo Institute of Technology (December 2022), the University of Pisa (scheduled for May 2023), and the Chongqing Normal University (scheduled for August 2023). His research trend continues along the lines of the previous year, with additional focus on linesearch-free methods for convex optimization [2] and (convex) optimization for power-grid expansion planning [4]. As in the past, a first inquiry in the convex realm is meant to serve as foundation for potential nonconvex extensions, ultimate target of the proposal.[1] Z Wang, AT, H Ou, and X Wang. A mirror inertial forward-reflected-backward splitting: Global convergence and linesearch extension beyond convexity and Lipschitz smoothness, arXiv:2212.01504, 2022[2] P Latafat, AT, L Stella, and P Patrinos. Adaptive proximal algorithms for convex optimization under local Lipschitz continuity of the gradient, arXiv:2301.04431, 2023[3] S Hardy, AT, K Yamamoto, H Ergun, and D Van Hertem. Optimal grid layouts for hybrid offshore assets in the North Sea under different market designs, arXiv:2301.00931, 2023

该候选人与不列颠哥伦比亚大学 (CA) 建立了新的合作关系，其中一名博士生在他的共同指导下进行了讨论，最终形成了预印本 [1]。此外，他还受邀到东京工业大学（计划于2022年12月）、比萨大学（计划于2023年5月）和重庆师范大学（计划于2023年8月）进行访问。他的研究趋势延续了前一年的路线，额外关注凸优化的无线搜索方法 [2] 和电网扩展规划的（凸）优化 [4]。与过去一样，凸领域的首次查询旨在作为潜在非凸扩展的基础，这是该提案的最终目标。 [1] Z Wang、AT、H Ou 和 X Wang。镜像惯性前向反射后向分裂：超越凸性和 Lipschitz 平滑度的全局收敛和线搜索扩展，arXiv:2212.01504, 2022[2] P Latafat、AT、L Stella 和 P Patrinos。局部 Lipschitz 梯度连续性下凸优化的自适应近端算法，arXiv:2301.04431, 2023[3] S Hardy、AT、K Yamamoto、H Ergun 和 D Van Hertem。不同市场设计下北海混合海上资产的最佳电网布局，arXiv:2301.00931，2023

项目成果

A Bregman Forward-Backward Linesearch Algorithm for Nonconvex Composite Optimization: Superlinear Convergence to Nonisolated Local Minima

• DOI: 10.1137/19m1264783
• 发表时间: 2019-05
• 期刊: SIAM J. Optim.
• 作者: Masoud Ahookhosh;Andreas Themelis;Panagiotis Patrinos

Douglas-Peucker piecewise affine approximation of an optimal fuel consumption problem to apply PANOC

应用 PANOC 的最佳燃油消耗问题的 Douglas-Peucker 分段仿射近似

• DOI: 10.23919/siceiscs54350.2022.9754372
• 发表时间: 2022
• 期刊: SICE International Symposium on Control Systems
• 作者: H. Ou;A. Themelis;T. Yuno and T. Kawabe
• 通讯作者: T. Yuno and T. Kawabe

UniBw Munich(ドイツ)

慕尼黑 UniBw（德国）

Flock navigation with dynamic hierarchy and subjective weights using nonlinear MPC

• DOI: 10.1109/ccta49430.2022.9966067
• 发表时间: 2022-02
• 期刊: 2022 IEEE Conference on Control Technology and Applications (CCTA)
• 作者: Aneek Nag;Shuo Huang;Andreas Themelis;Kaoru Yamamoto

inertia and relative smoothness in nonconvex minimization a case study on the forward-reflected-backward algorithm

非凸最小化中的惯性和相对平滑度——前向反射后向算法的案例研究

• 发表时间: 2022
• 作者: M. Ahookhosh;A. Themelis;and P. Patrinos;Andreas Themelis
• 通讯作者: Andreas Themelis