Prediction-free Online Dispatch of Remote Wind-Storage Plant Considering Transmission Congestion

Large-scale renewable energy generations rapidly develop in recent decades; however, the inherent uncertainty of renewable resources, as well as the insufficiency of transmission corridors, injures the operating economy of renewable plants. Energy storage is considered as a promising support in favor of exploiting the profitability of renewable energy generations. This paper investigates the real-time operation of a wind-storage system with the aim of profits maximization. Energy storage plays a role in shifting energy to improve the electricity sale incomes and reduce the curtailment cost in a transmission capacity-constrained and time-of-use price environment. Unlike other forecast-based rolling optimization methods, we employ the Lyapunov optimization theory to relax the time-coupling constraints of energy storage and achieve prediction-free online scheduling. The deduction based on Lyapunov optimization theory is given; the lemma that determines the critical penalty coefficient in the drift-plus-penalty problem is proved. In the online stage, it entails solving a small-scale mixed-integer linear program to obtain the scheduling strategy. The statistical optimality gap can be easily estimated. Case studies on a designed wind-storage system validate the performance of the proposed method.