Online Critical Load Restoration for PV-Storage Integrated Distribution Systems Considering Update of Forecast Uncertainty

Power systems are vulnerable to natural disasters and thus a proper critical load restoration strategy after black-outs is significant for the resilience enhancement of power systems. Current studies regarding the restoration of PV-storage integrated distribution systems cannot handle the scenario-based stochastic optimization with rolling updated forecast uncertainty of PV power outputs. To fill this gap, this paper proposes a systematic online critical load restoration scheme, based on a novel Gaussian mixture model-based scenario generation method, and the mathematical optimization is also formulated with the second order cone programming form. The case study to the IEEE 37-node test feeder shows the importance of forecast uncertainty update to the improvement of restoration rate. The whole scheme can provide guidance of emergency dispatch under main grid outages for distribution system operators.