Capacity And Location Optimization Of Photovoltaic And Energy Storage Based On Bidirectional Dynamic Reconfiguration And Cluster Division

With the continuous deepening of research on distributed renewable energy power generation and the encouragement of a series of favorable domestic policies,distributed power generation represented by photovoltaics(DPV)is showing a trend of rapid regional development.Large-scale,clustered distributed power generation is in parallel.The Internet will be the key development direction in the future.Energy storage system(ESS)can effectively alleviate the timing mismatch between source and load with its fast response,providing a solution for large-scale DPV grid connection.Due to the high cost of ESS at this stage,reasonable planning to increase economic benefits is also one of the research focuses.Aiming at the problem of the consumption of high-permeability DPV in the distribution network,a two-layer partition model is proposed,and based on the results of the two-layer partition,the DPV and ESS are selected and determined in stages,and its effect on improving the consumption of DPV is analyzed.To reduce the overall cost of the system.The thesis mainly does the following work:1)Seek an efficient,accurate and fast distribution network reconfiguration algorithm.In the optimization process of the two-layer model,frequent reconstruction will be involved.A heuristic network reconstruction method based on the random spanning tree method is proposed,which can quickly obtain the network reconstruction results,Compares the 33-node and 69-node calculation examples with other intelligent algorithms to verify their accuracy and rapidity,which can be regarded as an efficient and reliable means of network reconstruction in the complex simulation process.2)In order to study the relationship between photovoltaic consumption and costs,establish various cost models and analyze the relationship between ESS and planning costs under the condition of fixed DPV output,and obtain the best ESS operation strategy and the relationship between optimal configuration and DPV;then use DPV output as a variable to observe the change trend of planning costs,and then obtain the "DPV consumption-planning cost" curve under the fixed network topology.Based on this,the movement trend of the "DPV consumption-planning cost" curve was observed through three kinds of network loss change schemes to verify the feasibility of dynamic reconfiguration to increase photovoltaic consumption and reduce planning costs.3)Establish a two-layer model.The upper layer is divided into time periods as decision variables,and the lower layer is divided into clusters as decision variables.A two-layer improved GA embedded in the random spanning tree reconstruction method is proposed to solve the model and adjust the weights of the number of time periods,cost and the cluster structure index.Analyze the results of the two-layer division,find the appropriate weight distribution,and take the 33-node system as an example to give a specific time-sharing dynamic reconstruction strategy Compare the difference between the cluster division result and the traditional division method with the cluster division scheme,as well as the “DPV consumption-planning cost” relationship curve before and after dynamic reconstruction,and the overall DPV and ESS constant capacity results of the system ignoring the differences at the node level(first stage)).Verify the validity and reliability of the two-layer model,and provide a basis for the next step of node-level DPV and ESS site selection and capacity(the second stage).4)Carry out optical storage site selection and capacity determination based on the two-layer model,and propose an improved PSO that single-layered the "cluster-node" two-layer planning architecture,combining three reference schemes,analysis and comparison The results of DPV and ESS planning and configuration of each program,as well as system operation indicators such as network active power loss,node voltage level,and cluster operation performance evaluation indicators such as cluster selfbalance and energy penetration rate.Analyze the effect of the dynamic reconfiguration strategy on reducing the annual overall cost of the system and improving the DPV consumption level,compare the speed and efficiency of the cluster-based single-level planning algorithm and the two-level iterative algorithm,and further verify the effectiveness of the two-layer model for optimizing DPV and ESS planning configuration.