Adaptive time resolution rolling dispatch with high renewable penetration

To cope with the energy crisis and the environment issues caused by the fossil fuels, the renewable energy generation developed rapidly. However, the inherent characteristics of uncertainty bring about difficulty to the operation of power system. Rolling dispatch can use the updated forecast data to adjust unit planning and it is a good method to cope with the volatility of renewable energy. But with the dispatch time resolution be finer, the computation burden of dispatch agent increases significantly and the allowable time limit of each round rolling dispatch shrunk. This paper proposed a time sections clustering method, which could adaptively reduce the number of time sections that should be considered in the optimization model. In order to avoiding the interval of two successive dispatch horizon, this paper use an extended horizon to cover a first few hours of next day. To adopt the variable time resolution, a unified unit commitment model is proposed. The numerical experiment on a real provincial system in northwest China shows the effectiveness of the proposed model.