[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86087-en":3,"doc-seo-86087-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},86087,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Reinforcement Learning versus Optimization for Optimal Transmission Switching: A Comparative Study","Optimal Transmission Switching (OTS) targets lower generation costs by strategically opening transmission lines, but its mixed-integer linear program (MILP) becomes computationally difficult for large-scale power grids. Reinforcement learning (RL) can provide a faster alternative, yet many RL-based OTS methods use soft penalties that allow physical constraint violations. This study compares an RL framework with an MILP optimization approach for OTS using IEEE RTS-96 24-bus case studies, evaluating solution quality and feasibility under different switching budgets.","Reinforcement Learning versus Optimization for Optimal Transmission Switching:  \nA Comparative Study  \nIsrael Abiala and Yuanrui Sang Dept. of Electrical and Computer Engineering  \nUniversity of Massachusetts Amherst Amherst, MA, USA {iabiala,[ysang](ysang}@umass.edu)[}](ysang}@umass.edu)[@umass.edu](ysang}@umass.edu)  \nRachel (Ryan) M. Gerdes  \nDept. of Electrical and Computer Engineering  \nVirginia Tech Arlington, VA, United States [rgerdes@vt.edu](rgerdes@vt.edu)  \narXiv :2607 . 10948v1 [ ee ss . SY] 12 Jul 2026  \nAbstract—Optimal Transmission Switching (OTS) reduces generation cost by strategically opening transmission lines, but its mixed-integer linear program (MILP) formulation scales poorly for large-scale transmission networks. Reinforcement learning (RL) offers a computationally efficient alternative, but existing RL-based OTS approaches rely on soft penalties that permit physical constraint violations. This paper presents a comparison between an RL framework and an MILP-based optimization method for OTS. Case studies were carried out on the IEEE RTS- 96 24-bus system; results show that the agent was able to produce near-optimal solutions at low switching budgets and tended to yield suboptimal solutions at high switching budgets. However, the RL agent was able to generate feasible solutions two-to-three orders of magnitude faster than the optimization solver.  \nIndex Terms—Behavioral cloning, deep Reinforcement Learning, grid-enhancing technologies, optimal transmission switching, soft actor-critic  \nNOMENCLATURE  \nSets/Indices  \nB Set of buses  \nG Set of generators  \nGn Set of generators connected to bus n L Set of transmission lines  \nS Set of linearized generation segments  \nδinn Set of lines entering bus n δoutn Set of lines leaving bus n  \nParameters  \nCnlg No load cost of generator g  \nCsgeg,s Marginal cost of segment s of generator g Fmink Minimum thermal capacity of line k  \nFmaxk Maximum thermal capacity of line k Sbase Base MVA of the system xℓ Reactance of line k  \nbℓ Electrical susceptance of line k (= Sbase /xk ) Pming Minimum generation of generator g  \nPmaxg Maximum generation of generator gdn Demand at bus n  \nM Large number  \nC Upper limit on the number of open transmission  \nlines  \nPsg,esg Upper limit of segment s of generator g θminn Minimum voltage angle of bus n  \nθmaxn Maximum voltage angle of bus n  \nVariables  \nzkor close Dn  \nFk  \nθn  \nPsg,esg  \nPg  \nBinary variable indicating whether line k is open  \nLoad demand at bus n Active power flow of line k Voltage angle of bus n  \nPower generation of segment s of generator g Total generation by generator g  \nI. INTRODUCTION  \nAs transmission congestion increases as a result of the integration of renewable energy sources into the power grid, transmission switching is used to mitigate the congestion and reduce power system operating costs [1] . According to the National Transmission Needs Study report released in 2023, implementing grid enhancing technologies like transmission switching can increase the effective utilization of transmission lines by 16% and reduce line overloading by at least 40%[2] . Several studies have investigated the optimal transmission switching (OTS) problem for different test cases. Traditionally, it is formulated as a DC optimal power flow (DCOPF)-based optimization model, which determines the optimal network topology and generator dispatch to meet a certain demand. In this case, the optimization problem is a mixed Integer problem, which is NP-hard but can be effectively solved for small-scale systems and significantly reduce transmission congestion, and,  \nconsequently, generation dispatch cost [3] . To study the longterm impact of transmission switching, the OTS problem can be integrated into unit commitment models considering N-1 security [4] . The OTS problem can also be integrated into along-term planning model, so that transmission expansion can be done considering the flexibility that transmission switching can offer","cbCaiakxu7qUd0mJ","https://ap.wps.com/l/cbCaiakxu7qUd0mJ","pdf",746915,4,1,6,"English","en",105,"# Abstract\n# Introduction\n## Background and motivation for OTS\n## Traditional MILP-based formulations\n## Prior algorithmic approaches\n## Machine learning and RL for OTS","[{\"question\":\"Why is Optimal Transmission Switching (OTS) computationally challenging with MILP?\",\"answer\":\"OTS is commonly formulated as a DCOPF-based mixed-integer problem. While effective for small systems, the MILP structure scales poorly and becomes hard to solve for large-scale transmission networks.\"},{\"question\":\"What limitation exists in many existing RL-based OTS approaches?\",\"answer\":\"Existing RL methods often rely on soft penalties, which can permit violations of physical constraints instead of strictly enforcing them.\"},{\"question\":\"What do the comparative case studies show about RL versus optimization?\",\"answer\":\"On the IEEE RTS-96 24-bus system, the RL agent produces near-optimal solutions at low switching budgets, but can be suboptimal at high switching budgets. However, it generates feasible solutions two-to-three orders of magnitude faster than the optimization solver.\"}]",1784208420,15,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"reinforcement-learning-versus-optimization-for-optimal-transmission-switching-a-comparative-study","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":20},"https://docshare.wps.com/document/reinforcement-learning-versus-optimization-for-optimal-transmission-switching-a-comparative-study/86087/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-26","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is Optimal Transmission Switching (OTS) computationally challenging with MILP?","Question",{"text":75,"@type":76},"OTS is commonly formulated as a DCOPF-based mixed-integer problem. While effective for small systems, the MILP structure scales poorly and becomes hard to solve for large-scale transmission networks.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What limitation exists in many existing RL-based OTS approaches?",{"text":80,"@type":76},"Existing RL methods often rely on soft penalties, which can permit violations of physical constraints instead of strictly enforcing them.",{"name":82,"@type":73,"acceptedAnswer":83},"What do the comparative case studies show about RL versus optimization?",{"text":84,"@type":76},"On the IEEE RTS-96 24-bus system, the RL agent produces near-optimal solutions at low switching budgets, but can be suboptimal at high switching budgets. However, it generates feasible solutions two-to-three orders of magnitude faster than the optimization solver.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,127,130,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":22,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]