Real-time control method and performance evaluation of pumped storage system based on adaptive moving average channel strategy

To address the challenges of delayed response and insufficient regulation accuracy in pumped storage systems under high-penetration renewable energy scenarios, this paper proposes a real-time control method based on an Adaptive Moving Average (AMA) channel strategy. The method treats water head as the core control variable and constructs a dynamic regulation channel that enables real-time trend tracking and boundary identification. When the head state deviates beyond adaptive channel limits, the system triggers output adjustment procedures. Compared to conventional PI and expert-rule-based strategies, the proposed method introduces a self-tuning mechanism and fault-tolerant recovery logic, enhancing state awareness and boundary-driven control responsiveness. The AMA channel framework—adapted from financial technical analysis—offers superior flexibility and trend adaptability, enabling precise regulation under volatile grid conditions. Simulation studies across multiple scenarios, including frequency disturbance response, AGC signal tracking, and head fluctuation interference, demonstrate that the proposed strategy significantly reduces regulation deviation (RMSE: 8.95 vs. 21.42), improves response speed, and maintains economic dispatch performance. Ablation experiments further confirm its robustness under parameter failure conditions. These results validate the method’s engineering applicability and its potential for intelligent control deployment in complex power systems.