Improved Successive Branch Reduction for Stochastic Distribution Network Reconfiguration

Abstract: We propose an improved successive branch reduction (SBR) method to solve stochastic distribution network reconfiguration (SDNR), a mixed-integer program that is known to be computationally challenging. First, for a special distribution network with a single redundant branch, we propose an improved design for a one-stage SBR algorithm in the literature to incorporate uncertain renewable generations and loads. Based on solving stochastic optimal power flow, the improved algorithm identifies and searches through a small set of candidate branches, from which it determines the optimal branch to open and obtains a radial network with the minimum expected operational cost. Then, for a general network with multiple redundant branches, we design a heuristic two-stage SBR algorithm based on a close-and-open procedure that iteratively runs the proposed one-stage SBR algorithm. Numerical results on the IEEE 33-bus and 123-bus distribution network models verify the proposed method in terms of optimality and computational efficiency.