Large-Scale Maintenance and Unit Commitment: A Decentralized Subgradient Approach

Unit Commitment (UC) is a fundamental problem in power system operations. When coupled with generation maintenance, the joint optimization poses significant computational challenges due to coupling constraints linking maintenance and UC decisions. Obviously, these grow size of network. With introduction sensors for monitoring generator health condition-based maintenance(CBM), have been magnified. ADMM-based decentralized methods shown promise solving large-scale problems, especially vertically integrated systems. However, their current form, fail deliver similar performance scalability when considering CBM problem. This paper provides novel framework large-scale, problems. Our approach relies on use subgradient method temporally decouple various subproblems formulation along horizon. By effectively utilizing multithreading, our delivers superior eliminates need move sensor data thereby alleviating privacy security concerns. Using experiments large scale test cases, we show that can provide speedup upto 50x as compared state art benchmarks without compromising solution quality.