Online Optimization using Distributed Algorithm for Power System Stability

The increasing integration of renewable energy sources has introduced significant uncertainty into power networks, which impacts the stability and of the system. This study presents an approach to strengthen power network stability by online optimization based on a distributed algorithm. The proposed framework employs online convex optimization that allows real-time adjustments of power flows based on small updates in cost parameters. In the power network with buses representing both generators and loads, each bus has access to only local information and also interacts with adjacent buses to optimize the global performance of the network. Our distributed primal-dual algorithm is designed to minimize static regret and constraint violations, ensuring that the network adapts efficiently to dynamic conditions. Methods such as Model Predictive Control (MPC), feedback optimization and stochastic approaches have also been explored. These methods, typically rely on prior knowledge of cost functions or probabilistic models of uncertainties. Extensive simulations on the IEEE-14 bus system demonstrate the superiority of our algorithm in maintaining stability and optimizing power flows under varying conditions. A number of experiments on the IEEE-14 bus system is implemented and simulation results are analyzed which shows that our algorithm outperforms others in terms of both stability and optimization of power flow, which indicates its capability to facilitate the transition towards the integration of renewable energy sources into current complex power networks.