Event-triggered Based Rolling Optimization in a Microgrid under Uncertainties

This paper introduces a scheduling method based on event-triggered rolling optimization to realize operation optimization of a microgrid. An event occurs when deviation of real-time values and predictive values exceed a predetermined range. Under this approach, an aperiodic rolling window is utilized to deal with immense volatility of a microgrid caused by weather condition, demand profile and time-of-use power price. Besides, in pursuit of accuracy of solutions, short time interval also increases computational burden. Continuous scheduling periods are incorporated into a rolling window to improve computational efficiency if there are no events. Simulation results show that the event- triggered rolling optimization method reduces the total operation cost and the number of iterations of operation under uncertainties. Moreover, the proposed method makes a trade-off between effectiveness of the algorithm and optimality of the scheduling.