Power Reference Tracking in Wind Farms Through Distributive Model Predictive Control

With the increasing use of wind energy as a power source, new challenges arise. One of these challenges is maintaining a stable power grid. On the power grid, the amount of energy generated and consumed should be in balance. As wind power has an unpredictable and fluctuating nature, it is presumable that its increasing use makes it problematic to maintain this balance. As most wind energy is generated at sites consisting of multiple wind turbines, called wind farms, this thesis focuses on these sites to overcome this problem. In this thesis, a distributed model predictive control (MPC) strategy is introduced that not only stabilizes the power produced by wind farms, but also creates the possibility to perform power reference tracking with wind farms. With power reference tracking, it is possible for grid operators to adapt the power production to a change in the power demand and to counteract fluctuations introduced by other power generators. For this MPC, in this thesis, a new low-fidelity control-oriented wind farm model is developed. In this control model, the wake dynamics are taken into account. Wakes are areas downwind from turbines with decreased wind speed and increased turbulence. These wakes cause the wind turbines within a wind farm to influence each other. It is envisioned that taking these effects into account, will benefit the tracking quality of the controller. Because with centralized MPC it becomes problematic to provide real-time control for large wind farms due to the large order of the controller model required for such wind farms, the controller proposed in this thesis uses distributed control. With distributed control, the central control problem is divided into smaller local control problems that will be solved in parallel on local controllers. This makes it possible to also solve large complex control problems in real-time.Mechanical Engineering | Systems and Contro