Research on load transfer scheme optimization with probabilistic graphical computing based distribution network operation mode verification

Abstract

In this paper, an optimization method for distribution system load transfer scheme with consideration of the operation mode verification is proposed. The method aims to recover the distribution system from the largescale power outage caused by complete shutdown of the substations. Firstly, a physical-probabilistic-network (PPN) model is introduced to estimate the operation mode of the distribution network by using belief propagation algorithm to recognize the states of switches. Then, a two-stages optimal transfer scheme which takes both the shortest recovery time and maximum load recovery rate as objectives is brought forward to determine alternative feeders for the outrage loads. The substations with high probability of overload after the first-stage of optimization are identified as the “weak links” and the second-stage optimization strategy which considers the capacity constraints is applied to further transfer the load from the “weak links” to their neighbours. Thus, the overload chances for the “weak links” are reduced and the overall load recovery rate is improved. Case study results prove that the verification of the distribution network operation mode can help improve the accuracy of the two-stages optimal load transfer scheme. Meanwhile, the proposed algorithm can effectively reduce the recovery time of the outage loads.