基于高弹性电网的虚拟电厂短期负荷预测

Short-Term Load Forecasting for Virtual Power Plant Based on Highly Resilient Grid

虚拟电厂是电厂智能发展的结果,是能源互联网的终极组态,它将需求侧资源和分布式能源与传统电网进行融合,共同满足社会发展对电力的需要。目前,虚拟电厂技术大部分应用在分布式能源的调度优化,而对用户侧需求响应的研究较少,且存在虚拟电厂需求侧资源和分布式能源的电能质量与可靠性问题。为解决上述问题,本文借助智能化技术进行调控,实现与传统电网的稳定融合,并对虚拟电厂接入的需求侧资源和分布式能源通过聚类、关联度分析、决策算法、回归算法等进行负荷预测和分析,以支撑虚拟电厂需求响应、清洁能源消纳等业务。仿真实验表明该预测模型可实现需求侧负荷、分布式电源出力的精准预测,支撑虚拟电厂实现分层分区精准调控及其他业务。

Virtual power plant(VPP)is the result of the intelligent development of power plants and the ultimate configuration of the energy Internet.It integrates demand-side resources and distributed energy with traditional power grids to jointly meet the needs of social development for electricity.At present,the virtual power plant technology is usually used in the dispatching optimization of distributed energy,while the research on user-side demand response is less,and there are power quality and reliability problems of virtual power plant demand-side resources and distributed energy.In order to solve above problems,this paper uses intelligent technology to control and achieve integration with the traditional power grid.Through using cluster analysis,correlation analysis,decision-making algorithm,regression and other algorithms to forecast and analyze load for the demand-side resources and distributed energy connected to the virtual power plant,which can support virtual power plant demand response,clean energy consumption and other businesses.Simulation experiment shows that the prediction model can accurately predict demand-side load and distributed energy power output,which can support the virtual power plant to achieve precise regulation and other services in layers and regions.