Uncertainty Characterization Integrated Stochastic Model Predictive Control for Shared Energy Storage Planning

With the significant increase in photovoltaic (PV) generation, the uncertainty faced by power grid operation has notably escalated, and the inherent uncertainty of PV generation itself is challenging to accurately characterize. Although energy storage technology can mitigate the impact of PV uncertainty by shifting energy demand, the configuration of energy storage is confronted with issues due to the high investment and long payback. In view of this, we presents a shared energy storage planning for interconnected regional grids based on uncertainty characterization integrated stochastic model predictive control. Firstly, we employ a Gaussian mixture model to describe the output curve of PV generation and utilize Latin hypercube sampling to model the uncertainty. Secondly, we establish the stochastic optimization for planning of shared energy storage. Subsequently, to cope with the impact of intra-day forecasting errors, we establish a shared energy storage operational model based on stochastic model predictive control. The case study results demonstrate that the effectiveness of the proposed method in describing the uncertainty of the PV generation.