Parallel Load Restoration of Distribution Network via A Shortest-path-based Grid Partitioning Approach

After the power system fails, it is necessary to restore the power supply to the important loads as soon as possible. For the distribution network with multiple black-start units, dividing the distribution network into multiple subsystems for parallel restoration can speed up the speed of distribution network restoration. This paper proposes a parallel restoration model for distribution networks via a shortest-path-based portioning approach. A distribution network partitioning method based on the shortest path algorithm is introduced, which filters out the unimportant load nodes on the edge by means of power verification. Based on the obtained partition scheme, a parallel restoration model for distribution networks is established aiming at maximizing the cumulative weighted load recovery amount. Restoration constraints of multiple components are comprehensively incorporated, including the power flow equation constraints, DG output constraints, load constraints, node voltage constraints, line power constraints, restoration sequence constraints, network topology constraints and initial constraints. The linear programming solver ILOG CPLEX is used to solve the proposed optimization model, so as to obtain the load recovery optimization scheme in each partition. The effectiveness of the proposed model and strategy is verified on a modified IEEE-69 bus distribution system.