Probabilistic Forecasting Based Adaptive Power Smoothing Framework for Hybrid Wind-Storage Systems

Weiting Qian,Can Wan,Changfei Zhao,Zhiyi Li

DOI: https://doi.org/10.1109/appeec48164.2020.9220419

2020-01-01

Abstract:With the increasing penetration of wind power, the inherent intermittency and fluctuation of wind power generation bring severe challenges to the secure operations of power systems. Hybrid wind-storage systems equip a wind farm with multiple energy storage devices and are capable of smoothing wind power fluctuations effectively. This paper proposes a novel probabilistic forecasting based adaptive power smoothing framework for hybrid wind-storage systems. To begin with, probabilistic wind power forecasting is combined with multivariate copula function to generate the temporally correlated wind power scenario set. Then, an adaptive variational mode decomposition technique is proposed to decompose the wind power scenarios into various mode components, which are used to calculate the target grid-connected wind power and the target charging/discharging power of different energy storage devices. With these target power allocations, a stochastic rolling optimization model is established to obtain the schedule plan for hybrid wind-storage systems for the purpose of smoothing power fluctuations and enhancing system operational economy. The effectiveness of the proposed adaptive power smoothing framework and the feasibility of utilizing hybrid energy storage system to smooth wind power fluctuations are verified by case studies based on actual data from a Danish wind farm.