Research On Distributed Smart Grid Automation Key Technology And Its Engineering Application

In order to achieve the national strategic target of "carbon emission peak and carbon neutralization",constructing a renewable energy-dominated novel power system is the reasonable way for the power industry to promote its own carbon emission reduction and support the carbon emission reduction of the whole society.The power system plays an important role in promoting the traditional energy saving and consuming the green energy power.Driven by continuous policy initiatives and unprecedented business development opportunities,large-scale centralized new energy power generation projects are in full swing.At the same time,a large number of small-scale and distributed energy power generation have emerged around the medium and low voltage distribution system.A new form of distributed smart grid for the assembly and management of distributed generation-grid-load-storage resources emerges as the times require.Distributed smart grid aggregates the energy resources of differentiated operating entity and random operation characteristics.Its on-grid operation and management pose a great challenge to the power system.The traditional energy power system vertical regulation architecture model can not meet the needs of large-scale distributed smart grid on-grid operation.Under the background of the gradual increase in the penetration of distributed generation,the traditional power system’s ’plug and forget’ response strategy to distributed energy power generation has seriously increased the risk of safe and economic operation of the power system.The distributed smart grid has become an important topic in the research of energy power system at the current stage.However,the information accessing mode,technology and operation strategy and methods around the distributed smart grid are still lack of systematic solutions.In the meanwhile,the distributed smart grid is considered as the gathering point of the ’information flow’,’service flow’ and ’energy flow’ of the novel power system.During the construction process,the distributed smart grid faces technical constraints at the prospect of communication,computation and function.Therefore,this paper focuses on the distributed smart grid system technical issues,including architecture,communication constraints,computation constraints,management strategies,and carries out the following targeted research:(1)The automation architecture considering to the requirements of multi-entity operating objectives,multi-regulation scale coordination and multi-aggregation mode overall planning for distributed smart grid is discussed.An automation architecture with the characteristics of’openness’,’integration’ and ’flexibility’ is designed to meet the application needs of direct control and non-direct control scenarios for distributed smart grid,and the secure access mode between the distributed smart grid and the traditional electric power automation system is also provided.(2)The extended standardized information modeling method based on IEC61850 and a heterogeneous communication networking scheme for distributed smart grid are proposed.Based on IEC 61850,the standardized information modeling of distributed generation-grid-load-storage resource objects and new operation services in distributed smart grid is carried out.In this way,the general description of the equipment attributes of the new objects and the detailed description of the functional attributes of novel emerging services are obtained,thus providing the northbound standard information accessing scheme.Based on the methods of network performance analysis and comprehensive evaluation considering subjective and objective weights,the research on the heterogeneous communication networking scheme within the distributed smart grid is carried out to meet the communication performance requirements of multi-service information interaction.(3)The computation-load simulation analysis and service unloading method based on Edge Cloud Sim for distributed smart grid is proposed.The method of service unloading considering the computation resource configuration for cloud-edge computation coordination in distributed smart grid is studied.Focusing on the distributed smart grid service computation-load decomposition and time-sequence processing model,the cloud-edge resource collaboration computation-load modeling,cloud-edge resource modeling,and resource sequence curve modeling by Edge Cloud Sim are established to provide the quantitative analysis method for distributed smart grid cloud-edge computation-load optimal uploading.Besides,the option boundary of unloading strategy for specific service between cloud-side and edge-side under different computation resource configuration scenarios are obtained.(4)The core operation algorithm based on Actor-Critic for distributed smart grid is proposed.Based on stochastic model description by Markov chain and Actor-Critic reinforcement learning algorithm,an intelligent decision model for distributed smart grid is established.The economic optimization objectives is described for the integrated operation of direct control and non-direct control function for generation-grid-load-storage regulation in distributed smart grid.In this way,the intelligent decision-making basis suitable for multi-agent participation and multi control means coordination is formed.(5)The distributed smart grid generation-grid-load-storage intelligent integration system is developed.Relying on Guangdong Provincial Key R&D Program and several science and technology projects of China Southern Power Grid,a platform base for the Internet of things with the integration of generation-grid-load-storage is constructed,on which the implementation of direct control and non-direct control functions of distributed smart grid are carried out.The pilot projects provide the engineering implementation model for the comprehensive construction of distributed smart grid.