Hongtao Mao, Yifeng Wang, Bin Dong, Yangyang Miao, Wu Ma, Jun Wang
{"title":"LLM-Enhanced PSO for ECU Configuration in Wireless-Supported Distribution Network Self-Restoration","authors":"Hongtao Mao, Yifeng Wang, Bin Dong, Yangyang Miao, Wu Ma, Jun Wang","doi":"10.1002/itl2.70135","DOIUrl":null,"url":null,"abstract":"<div>\n \n <p>Rapid fault self-restoration in complex power distribution networks is crucial. Edge Computing Units (ECUs) offer decentralized control, but their optimal configuration is challenging. This paper proposes a Large Language Model (LLM) enhanced Particle Swarm Optimization (PSO) framework for ECU configuration, explicitly considering advanced wireless communication (e.g., 5G/6G, LoRa) characteristics. The LLM aids in intelligent population initialization and adaptive particle guidance within PSO. This approach aims to optimize ECU placement and inter-ECU data exchange. Simulations on IEEE test systems show that the LLM-enhanced PSO significantly improves ECU configurations, reduces communication delays, and enhances self-restoration performance, thereby bolstering smart grid resilience.</p>\n </div>","PeriodicalId":100725,"journal":{"name":"Internet Technology Letters","volume":"8 5","pages":""},"PeriodicalIF":0.5000,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Internet Technology Letters","FirstCategoryId":"1085","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/itl2.70135","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"TELECOMMUNICATIONS","Score":null,"Total":0}
引用次数: 0
Abstract
Rapid fault self-restoration in complex power distribution networks is crucial. Edge Computing Units (ECUs) offer decentralized control, but their optimal configuration is challenging. This paper proposes a Large Language Model (LLM) enhanced Particle Swarm Optimization (PSO) framework for ECU configuration, explicitly considering advanced wireless communication (e.g., 5G/6G, LoRa) characteristics. The LLM aids in intelligent population initialization and adaptive particle guidance within PSO. This approach aims to optimize ECU placement and inter-ECU data exchange. Simulations on IEEE test systems show that the LLM-enhanced PSO significantly improves ECU configurations, reduces communication delays, and enhances self-restoration performance, thereby bolstering smart grid resilience.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
