Stochastic Dynamic Optimal Power Flow Integrated with Wind Energy Using Generalized Dynamic Factor Model

Integrating large wind power to power grid complicates the system operation due to the intermittency of wind. To address the challenges associated with wind power this paper formulates a stochastic dynamic optimal power flow to consider the uncertainty of wind power forecast in the power system operation. The goal of the system operation is to minimize the expected operational cost. The uncertainty of wind power forecast can be represented by a set of scenarios. For accurate forecasting, the generalized dynamic factor model (GDFM) is proposed to synthesize the forecasted wind power outputs while it preserves the correlation structure among wind power outputs from closely located wind farms under similar weather condition. In this paper, the GDFM was developed by the raw data of 94 wind farms in Texas and we use the model to predict hourly wind power outputs for 24 hours. This work also emphasizes on AC optimal power flow and is solved by a novel heuristic method, artificial bee colony (ABC). Such method is chosen because of its effective nature of handling hard constraints. Case studies are conducted on modified IEEE-30 system.