An Efficient and Accurate Distribution Service Restoration Algorithm Based on Stochastic Programming

To handle uncertainties from renewable energy sources (RESs) in distribution service restoration (DSR), an efficient and accurate DSR algorithm based on stochastic programing (SP) is proposed. Firstly, based on the scenarios, the model aiming at maximizing the amount of restored critical loads while minimizing the switching times of switches is established. Then the linearization technology is applied to reformulated the mixed integer nonlinear programming (MINLP) into a mix-ed linear programming (MILP). Finally, the spectral clustering method is used to make the MILP tractable by decrease the scale of the problem. Numerical simulations conducted on the modified IEEE 69-bus distribution network demonstrate that the proposed algorithm can better accommodate uncertainties in an efficient and accurate way.