Distributed Network Expansion Planning With Ditribution Generation Considering Time Sequence Characteristics Between Source And Load

With the continuous development of smart grid,the application of distributed generation(DG)has been growing steadily,and the demand for stability of power grid is getting higher and higher with the gradual expansion of the system and the gradual increase of load.Reasonable access to distributed generation in distribution network can reduce network loss and improve network stability.However,inappropriate location and capacitance can improve network loss,reduce voltage quality,and affect relay protection actions.In addition,the fluctuation of wind and solar output may increase the fluctuation of node voltage,and even affect the normal operation of power system.Therefore,reasonable optimization schemes should be adopted in the planning of distributed network with DG so as to achieve optimal results.In this paper,DG is planned on the basis of no new load nodes in distribution network.Variarional auto-encoder is used to learn the statistical rules of historical data,and a distribution model is obtained.The newly generated data are clustered to generate more realistic typical day scenes with wind-light-load time series characteristics.In the process of power flow calculation,the output of DG at different time and the corresponding load increase the computational complexity.When selecting DG-installable nodes,the index vector method taking into account the time sequence correlation between sorce and load can reduce the difficulty of optimization.Based on the time series characteristics of DG output and load,the artificial fish swarm algorithm is used to determine the planning scheme.When additional load occurs in the distribution network,the distribution network with distributed generation is expanded and planned.The initial scheme of distribution network expansion is determined by the root node fusion method.Then the infeasible solutions(islands,chains and rings)are repaired,and finally,a feasible set of grid schemes is obtained.The feasible grid scheme set and the location capacity of DG are coded uniformly.In order to reasonably simulate the whole process of distribution network operation,considering that the planning model involves many factors,such as technology and economy,and aiming at the lowest cost of network loss,the lowest cost of purchasing electricity from the main network,the lowest cost of distributed generation and the lowest cost of line investment and operation,the artificial fish swarm algorithm is used to solve the optimal distribution network planning scheme with distributed generation.Taking an actual 35-bus distribution network and 49-bus distribution network as an example,the planning of DG,distribution network expansion planning and their collaborative planning are simulated and analyzed respectively.The impact of different planning schemes on distribution network economy is also compared.It is verified that the planning model and new method constructed in this paper can effectively improve planning efficiency and reliability and economy of distribution network operation.