Research On Battery Energy Storage Capacity Configuration Of Virtual Power Plant Considering Classical Scene Sets

Most technologies that rely on natural energy have developed rapidly,leading to the massive consumption of fossil fuels.In this regard,many solutions have been proposed to alleviate phenomena such as air pollution and global warming.The concept of virtual power plant was born.Virtual power plant(VPP)can provide an effective means to manage distributed generation(DG)when distributed energy is growing rapidly.At present,there is no authoritative definition of VPP.The concept accepted by most people is that it uses advanced control,communication and computing technologies to aggregate different types of distributed energy sources in a distributed network and further distribute energy so that these DGs can be integrated Operation,that is,convenient management,can also effectively alleviate the fluctuation of distributed energy.In this concept,different distributed energy resources(DER)are linked together through a communication infrastructure,and operate as a single power plant in an aggregated manner.This method can take advantage of the synergy between different technologies and effectively trade aggregated feeds in the electricity market and the balanced electricity market.However,most of the current research on VPP focuses on dispatching operation,but there are only a handful of researches on VPP energy storage capacity allocation.In this context,this article will study the battery energy storage capacity allocation of virtual power plants.First,the characteristics of the virtual power plant and its frame structure are systematically analyzed.Through theoretical analysis,the mathematical model of wind power and photovoltaic power generation is established,the required modules of the virtual power plant are proposed,and the basic model of the virtual power plant is constructed.At the same time,by comparing Monte Carlo sampling and Latin hypercube sampling,Latin hypercube sampling is selected for stratified sampling of distributed power sources to ensure the uniformity of the sampled samples and form a scene set of scenery,thereby turning uncertainty into certainty.Then select K-means clustering and apply it to the scene set to filter out a large number of duplicate or similar data in the original data,thereby forming a classic scene set.Secondly,taking the maximum net income of the virtual power plant as the objective function,considering the impact of the time-of-use electricity price on the virtual power plant,through the energy interaction between the virtual power plant and the grid,the optimal configuration of energy storage capacity and the net income of the virtual power plant under this configuration are obtained by solving.Finally,considering that traditional particle swarm algorithm is easy to fall into a local optimal solution,we adopt a particle swarm algorithm with adaptive weights.At the same time,in order to strengthen the algorithm’s global search capabilities,the concept of hybridization in genetic algorithms is added to the particle swarm algorithm.Then,a virtual power plant including wind power,photovoltaic,gas turbine and energy storage is used as an example to analyze the calculation example,and the optimal configuration of the battery energy storage capacity of the virtual power plant is obtained,and the feasibility of the method is obtained by comparison.