Nonlinear model predictive control-based active power dispatch strategy for wind power plant considering dynamic wake effect

Abstract

Inappropriate active power dispatch strategies can diminish the active power support capability (APSC) of the wind power plant (WPP) due to the wake effect. In existing strategies, the WPP’s APSC and the time delay of wake effect are not thoroughly investigated. This paper proposes an active power dispatch strategy to increase the WPP’s APSC while tracking the power reference. First, WPP’s temporary and sustained APSCs are analyzed considering dynamic wake effect. Correspondingly, the total stored kinetic energy and increasable power of WPP are introduced as evaluation indicators of the capabilities. Furthermore, a bi-level control framework is proposed to accomplish the dispatch task in different timescales. Aiming at improving WPP’s temporary and sustained APSCs comprehensively, the operating status references of each WT, including the power reference and pitch angle, are optimized in the WPP layer. Taking the time delay and nonlinearity of wake effect into account, the receding horizon optimization is solved in a finite time domain using nonlinear model predictive control (NMPC). During the sampling interval of the NMPC, a coordinated control in the WT layer is developed to modify the active power reference dispatched by the WPP layer in case the wind speed fluctuates. Finally, extensive case studies show that the WPP’s temporary and sustained APSCs are improved with the proposed strategy in different wind conditions and power references.