Research On Demand Side Resource Cluster Management And Control Method For Edge Smart Station Area

Under the guidance of the “Carbon Peaking and Carbon Neutrality” target,China is accelerating the construction of a new power system with new energy as the main body.At the same time,the fluctuation and intermittentity of large-scale new energy after gridconnected make the system face enormous challenges.The development of demand response technology has mobilized the user-side flexible resources and enabled the extensive participation in the regulation of the power grid with massive response capabilities.It is an effective measure to promote the absorption of new energy,improve the flexibility of distribution network regulation and operation stability.However,the characteristics of differentiation,fragmentation,and scattered position of demand-side resources greatly hinder its development and utilization.The edge computing technology and the cloud-edgeend collaboration model based on it coincide with the high-proportion distribution characteristics of the resource,providing ideas for the requirements of the demand side to participate in control.This article conducts research on the multi-time-scale management and control method for the demand side resource cluster of the Edge Smart Distribution Station Area.The specifics are as follows:First,modeling and clustering the demand-side resources.The regulation model of various demand-side resources in the incentive response mode is established,and i Aiming at the problem that it is difficult to effectively integrate and utilize massive heterogeneous decentralized resources,a multi-dimensional feature clustering model of demand-side resources based on improved K-means is constructed.That is,by extracting the multidimensional regulation characteristics of various resources for clustering,the resources with similar space-time adjustable ability are formed into clusters and regulated by clustering centroid.Simulation results show that this model can effectively cluster resources with similar space-time adjustable potential.Secondly,under the cloud-edge-end architecture,the two-stage that day-ahead and day-in collaborative optimization scheduling model for demand side resources is constructed for Edge Smart Distribution Station Area.As a edge node,the Edge Smart Distribution Station Area integrates internal resources to present overall coordination functions to the cloud.After collecting data from each edge to the cloud,global optimization is performed,and the optimization results of Edge Smart Distribution Station Area regulation are distributed to each edge.Each Edge Smart Distribution Station Area ultimately completes the optimal management of demand side resources.Based on this cloud-edge-end architecture,a two-stage that day-ahead and day-in collaborative optimization management model for resource clusters in the Edge Smart Distribution Station Area is established,with the goal of minimizing the incentive and compensation costs paid by the station area,to achieve collaborative and optimal allocation of total adjustment tasks among internal resource clusters in the edge station area.The simulation verifies that the proposed strategies can count the differences and adjustable differences in resources,make full use of the refined adjustment information of the end-side resources,and reduce the total cost of incentive compensation in the Station Area.Finally,under the edge-collaborative architecture,a real-time distributed optimal scheduling strategy for demand-side resources in edge stations based on a consistency algorithm is constructed.In order to correct the real-time power deviation caused by the long-term scale prediction error to the distribution network,under the coordinated regulatory structure of the edges,using the incremental cost rate as a variable,in the upper layer a consistent iteration is performed to achieve the collaborative allocation of the total deviation correction tasks among the Edge Smart Distribution Station Areas of the station group;a real-time autonomous regulatory model is constructed in the Station Area to achieve real-time optimization and distribution of resource adjustment tasks in the station area.In the lower layer,a real-time autonomous model of the Station Areas is constructed to complete the real-time control of each cluster in the station area based on the coordinated allocation between stations.The simulation verifies the effectiveness and adaptability of the proposed upper layer distributed control strategy,and the lower layer real-time autonomous model fully utilizes the advantages of resource clusters.