Dynamic Equivalent Modeling with Physics-Data Fusion for Wind Farm Based on Characteristics of Multiple Spatio-Temporal Scales
A dynamic equivalent modeling method for wind farms is proposed.The core idea is to effectively integrate the physical characteristics analysis and data analysis of wind farms,to construct data features reflecting the transient characteristics of wind farms.And a clustering algorithm based on shape-based distance is proposed to address the transient fluctuations with variable characteristics of wind power,achieving clustering equivalence of wind turbines with similar dynamic behaviors within a wind farm.First,based on the multiple spatio-temporal scale characteristics of wind farms,the core factor set are proposed to characterize the dynamic characteristics of the wind farm.Then,considering the correlation between multiple core factors,the denoising autoencoder algorithm is employed to reduce dimensionality and eliminate redundant information to construct data features.In addition,a clustering algorithm based on the shape-based distance is introduced to achieve clustering of power generation units for wind farms.Finally,taking a certain wind farm as a case,the effectiveness of the proposed method is verified through simulation on the PSCAD/EMTDC platform.
equivalent modelingwind farmmultiple spatio-temporal scalesdenoising autoencoderclusteringphysics-data fusion model
万方数据
维普
