Optimisation of Wind-Solar-Storage Cooperative Operation Strategy for Virtual Power Plant based on Machine Learning
Aiming at the effective regulation of the operating load of virtual power plants with new energy access and the stable operation of the power system,a machine learning-based optimisation method for the cooperative operation strategy of virtual power plants with wind,light and storage is proposed.Taking the virtual power plant wind,light and storage cooperative operation as a benchmark,the probabilistic characteristic model is used to obtain the corresponding parameters of wind and light output,and then the source and load output characteristics of the virtual power plant under the wind,light and storage cooperative operation are portrayed;through the multi-dimensional filtering mode of machine learning,combined with the filtering algorithm to construct the feature evaluation function,in order to perceive the process of the wind,light and storage cooperative operation of the virtual power plant;based on the relevance of the objective function,and with reference to the attributes of the different features categories,using the feature grading theory to establish the grading structure;using evidential reasoning to rank the features,so as to determine the cooperative operation strategy of virtual power plant wind energy storage and complete the design of the optimisation method.The experimental results show that,in a power system test environment containing four groups of wind farms and photovoltaic systems,the application of the new co-optimisation strategy to different unit regulation modes can achieve the precise regulation of the co-operation operation indexes.This method can ensure the stable operation of wind and solar energy storage after integration into the power system and optimise the maximum output of thermal power units,which has practical application value.
NETL
NSTL
万方数据
