Predictive control of wind turbines with storage

The large-scale use of wind power generation continues to be hindered due to its intermittency. Among the potential solutions to this problem, the adoption of battery-based storage systems is widely seen as inevitable. The aim of this paper is to develop a real-time model-based optimisation approach for the coordinated control of a wind turbine equipped with battery storage. First, the mathematical model of the wind turbine-battery system is systematically reduced using singular perturbation theory. Then, the obtained reduced-order model is utilised in the control system development. The control system is devised using a real-time implementable version of model predictive control whereby the nonlinear dynamics are linearised at each sampling instant to simultaneously overcome the computational issues due to nonlinear optimisation and performance degradation issues due to linearisation at only one operating point. The effectiveness of the proposed controller in reducing wind intermittency through the optimal management of the battery storage is demonstrated using simulation studies involving real wind-data from an Australian wind farm.