Multi-Stage Stochastic Planning of Wind Generation Considering Decision-Dependent Uncertainty in Wind Power Curve

Abstract

Variability and uncertainty in wind generation brings new challenges to power system planning. In addition to inaccurate prediction of wind speed, recent studies have revealed that the actual wind turbine power curve (WTPC) is also uncertain before installation and operation of wind turbine (WT), which has unneglected impacts on future operations and should also be taken into consideration when make wind generation expansion plan. The uncertainty in actual WTPC is decision-dependent since its resolve time depends on previous investment decisions to acquire more accurate information of WTPC. Under this background, in this paper, we propose a multi-stage stochastic planning model for wind generation, considering both the exogenous uncertainty in load and wind speed prediction and the decision-dependent uncertainty in WTPC. First, the uncertainty modeling is presented and the decision-dependency of uncertainty in actual WTPC is illustrated. Then the multi-stage wind generation planning model is established with objectives of minimizing total expected cost in planning horizon, followed by physical constraints of power system and non-anticipativity constraints among scenarios. Case study demonstrated advantages of the proposed multi-stage stochastic planning of wind generation.