Accelerated Dynamic Event-Triggered Algorithm for Online Power System Optimization by Using Powerball Technique

Nowadays, online optimization for power systems has gained increasing attention due to many time-varying scenarios in practical applications. This paper proposes a novel feedback-based online algorithm for power system optimization problems by combining the powerball technique and a dynamic event-triggered mechanism, aiming to achieving a faster convergence rate and high communication efficiency simultaneously. In particular, a measurement-based prima-dual approach is utilized to solve the considered time-varying optimization problem. Taking the constrained resource issue, a more flexible triggering condition with a dynamic threshold is designed for dual variables to reduce communication burden. By further employing the powerball strategy, the primal-dual iteration is modified with an improved convergence rate. Finally, numerical experiments on test systems are conducted to demonstrate the effectiveness of the proposed online algorithm. It is shown that the proposed approach has superior convergence property especially in the initial process. Meanwhile, a better trade-off between the satisfactory system performance and communication resources consumption can be achieved.