基于区域间场景等值的配电网随机潮流加速算法OA北大核心CSTPCD

In the stochastic power flow algorithm of distribution network,the power injection needs a lot of simulation,which will lead to the problem of low computational efficiency.In recent years,scholars have made some achievements in the research of scene reduction,but the designed methods may still face the problem of low computational efficiency after being implemented in systems with too many nodes.Therefore,this paper proposes a stochastic power flow acceleration algorithm for distribution networks based on inter-regional scenario equivalence.Based on the existing mainstream scenario reduction method,the stochastic power flow calculation is further accelerated by adding scenarios that lead to the injection of the same boundary quantity in other regions. Firstly,based on the topology connection law of distribution network,the influence of power injection scenarios in each region on power flow in other regions is transformed into the influence on its boundary quantities,and the multi-regional power flow model that affects each other through boundary quantities is established.Secondly,an inter-regional scenario equivalence method is proposed,which can quickly equalize scenarios that cause the same power flow change in other regions.Then,a distributed solution algorithm for stochastic power flow is designed.Each region completes the distributed solution of stochastic power flow by exchanging boundary equivalent scenarios and combining local power injection information. According to this method,the IEEE 118 node test system is solved with partition and stochastic power flow.Simulation results show that the proposed distributed algorithm converges in all regions.In terms of calculation accuracy,the maximum errors of voltage and active power between the scene equivalent imitation method proposed in the main region and the Monte Carlo simulation method are 0.069% and 0.98% respectively,while the maximum errors of voltage and active power between the scene equivalent imitation method proposed in the sub-region and the Monte Carlo simulation method are 0.058% and 0.003% respectively.It can be considered that the calculation results are consistent.In terms of computational efficiency,the number of scenarios required by Monte Carlo simulation and distributed solution algorithms is 531 441 and 43 497,respectively.Compared with Monte Carlo simulation,the acceleration ratio of distributed solution method is 11.32,which effectively reduces the simulation scenarios required by stochastic power flow. Through the example analysis,the following conclusions can be drawn:(1)By merging the power injection scenarios that cause the same boundary quantity in other regions,the redundant information in the transmission process can be effectively reduced.(2)The proposed stochastic power flow distributed algorithm does not require centralized processing and unified calculation of the stochastic power flow injection samples of each node,but only needs to solve the stochastic power flow results through the power flow samples of each region and exchange the equivalent scenario information.(3)Compared with the existing centralized stochastic power flow algorithm,the distributed stochastic power flow algorithm can calculate more samples under the same time constraint.In the next step,the point estimation method can be combined with the proposed inter-zone scenario equivalence method to improve its calculation accuracy and enrich the distribution network region division method.