基于双层优化迭代模数并联双驱的电力系统负荷建模方法OA

To address the issues of insufficient modeling accuracy and weak generalization capability of traditional load modeling methods in emerging power systems,a novel power system load modeling method based on a dual-layer optimization iterative framework and a parallel model-data dual-driven architecture is proposed.By integrating physics-based and data-driven models through collaborative modeling,the limitations of single modeling approaches are overcome.First,the total load current is decomposed into two components based on an initial current decomposition strategy.Then,the load model is equivalently represented as a parallel structure with two branches,which are modeled by a physical model and a data-driven model,respectively.A dual-layer optimization iterative model is further employed to dynamically adjust the distribution of the input current and output power between the two branches,enabling mutual correction and collaborative optimization of the two models to construct a unified load model.Finally,the proposed method is validated using PMU measurement data from a power substation in Yunnan Province,China.The simulation results demonstrate that,compared with traditional methods,the proposed method achieves higher modeling accuracy and stronger generalization capability.