Research on Integrated Sensing, Communication, and Computing Power Scheduling Algorithms for Intelligent Sensing in Converter Stations

As a critical component of the power system, the operational stability and efficiency of the converter station directly affect the safety and reliability of the entire power system. The existing converter station management systems have issues such as insufficient utilization of computational resources and low scheduling efficiency, making it difficult to meet the growing power demands and complex operating environments. To address these issues, this paper proposes a computation power orchestration algorithm based on the converter station scenario, including a service layer, control layer, resource layer, and orchestration management layer. The computation power orchestration algorithm employs deep learning methods to predict and construct temporal computation power demand spaces within the scenario and utilizes reinforcement learning to converge on an optimal global computation power orchestration scheme.