Multi-time Scale Collaborative Optimal Dispatching for Regional Grid Integrated with Renewable Energy and Battery Energy Storage Based on GrowNet

With the development of renewable energy, Battery Energy Storage (BESS) has great potential in reducing the negative impact of renewable energy to modern power grid. In this paper, a forecasting framework based on gradient boosting neural network (GrowNet) is proposed, which can build a hierarchical deep neural network prediction model for the renewable energy. And a multi-time scale optimal dispatching framework based on improved model predictive control (MPC) is built for the system integrated with renewable energy and BESS, not only realizes the goal of lowest composite cost in day-ahead phase, but also reduces the fluctuation of tie-line power and the state of charge (SOC) of BESS, and tracks the planned power generation of renewable energy in intra-day phase. An improved genetic algorithm (GA) is used to solve the day-ahead optimal dispatching model, and the intra-day rolling optimal dispatching model is solved by quadratic programming combined with MPC. The simulation results show that the composite framework of forecasting and dispatching is practical and can achieve multiple optimization objectives.