🤖 AI Summary
To address the challenge that centralized real-time optimal load shedding (OLS) fails to meet millisecond-level response requirements during large-scale power system contingencies due to excessive computational and communication overhead, this paper proposes a decentralized machine learning framework. The method follows an “offline training, online autonomy” paradigm: each load center independently generates an optimal, power-flow-feasible load-shedding strategy using only local measurements via a distributed neural network. Innovatively integrating deep neural network modeling, distributed cooperative control, and explicit power-flow constraint embedding, the framework is validated on the IEEE 118-bus and Texas 2000-bus systems. In thousand-node-scale scenarios, it achieves decision latency under 10 ms and reduces communication load by over 90%, effectively mitigating cascading failures and significantly enhancing grid resilience and emergency robustness.
📝 Abstract
Prompt and effective corrective actions in response to unexpected contingencies are crucial for improving power system resilience and preventing cascading blackouts. The optimal load shedding (OLS) accounting for network limits has the potential to address the diverse system-wide impacts of contingency scenarios as compared to traditional local schemes. However, due to the fast cascading propagation of initial contingencies, real-time OLS solutions are challenging to attain in large systems with high computation and communication needs. In this paper, we propose a decentralized design that leverages offline training of a neural network (NN) model for individual load centers to autonomously construct the OLS solutions from locally available measurements. Our learning-for-OLS approach can greatly reduce the computation and communication needs during online emergency responses, thus preventing the cascading propagation of contingencies for enhanced power grid resilience. Numerical studies on both the IEEE 118-bus system and a synthetic Texas 2000-bus system have demonstrated the efficiency and effectiveness of our scalable OLS learning design for timely power system emergency operations.