Research On Distribution Network Cluster Partition And Voltage Control Based On Multiple Distributed Resources

With the growing problems of energy shortage and environmental pollution.Photovoltaic,wind and other decentralized resources to generate electricity with its abundant resources,widely distributed,clean and non-polluting advantages have received widespread attention and rapid development.The output of diversified decentralized resources is affected by natural conditions with intermittent,fluctuating and uncertain characteristics.Its large-scale and decentralized access has changed the one-way tidal distribution characteristics of the distribution network.making the power quality problems such as voltage crossing and power backward transmission of the distribution network more serious.It makes the power quality problems such as voltage crossing limit and power backward transmission in distribution network more serious.At the same time,the complexity of voltage control in the distribution network is greatly increased due to the many points of access to decentralized resources,which poses a new challenge to the voltage management of the distribution network.In this paper,the voltage control and optimal scheduling of the distribution network containing scaled multi-decentralized resources are studied:(1)A distributed resources cluster partition method that integrates cluster structure and function is proposed.The global mass density function is selected as a structural metric to measure the structural strength of the cluster.Ensuring that cluster segmentation meets its structural principles.The inconsistency coefficient and node affiliation function are selected as functional indicators.Where the inconsistency coefficient is used to describe the trend of distributed resources output within the cluster.The node affiliation function is used to determine the reasonableness of the cluster size.The selected metrics can take into account the overall performance of the cluster and reduce the complexity of subsequent voltage control and optimal scheduling.Then,the association partitioning algorithm is introduced through the theory of association partitioning to complete the cluster partitioning.The algorithm has the function of fast partition and realization of dynamic adjustment of distribution network under different time scales.(2)On the basis of cluster partition,a cluster voltage second-order cone optimization model with the goal of maximum distributed resources consumption within the cluster is established to achieve independent control of intra-cluster voltage.And the active power coordination interaction strategy for distributed iterative optimization among clusters is proposed to achieve coordinated optimization among clusters by using the characteristics of weak coupling between clusters.Fastly and effectively solve the overvoltage problem caused by the grid connection of distributed resources.(3)A distribution network voltage hierarchy control strategy combining global optimization and cluster control is proposed.Global optimization of scheduling level counting and decentralized resource reactive power regulation characteristics.By establishing a dual time-scale global optimization model,we realize active control and realtime optimization of the distribution network.Cluster control level reduces the node voltage deviation range based on the real-time voltage level of critical nodes.A cluster convergence control model based on power hierarchy is established to realize inter-cluster coordination optimization and intra-cluster power allocation.The proposed distribution network cluster voltage control method in this paper.Promote the orderly access and safe consumption of diversified and decentralized resources to ensure grid stability.At the same time,it provides theoretical support for solving the problems of voltage fluctuation and tidal backflow in distribution networks containing largescale multi-distributed resources.