Node Migration Based Optimized Network Partitioning Strategy for Power System Parallel Computation

The optimized partitioning for parallel computing of power network is researched. By means of Laplacian spectrum recursive bisection partitioning method, firstly the power network is partitioned into branch-cut form; then the branch-cut form is changed into node-split form. During the change process an optimized branch sort scheme is used to minimize border blocks; thus the time consumed in coordinative calculation is saved. By means of calculating the objective function of migrated nodes, the unbalance of partitioning is reduced. The testing calculations for partitioned network are conducted in IEEE 118-bus system, IEEE 162-bus system and IEEE 300-bus system respectively by parallel power flow algorithm. Calculation results show that the proposed optimized partitioning strategy can reduce bordered blocks effectively, so it is proved that the proposed optimized partitioning strategy is suitable for power system parallel computing.