基于哈默斯坦模型的分布式光伏发电系统建模与参数辨识  

Modeling and parameter identification of distributed photovoltaic power generation system based on Hammerstein model

作　　者：汤志锐 邱桂华 何引生 TANG Zhirui;QIU Guihua;HE Yinsheng(Guangdong Foshan Electric Power Bureau of China Southern Power Gird,Foshan,Guangdong 528000,China)

机构地区：[1]南方电网广东佛山供电局,广东佛山528000

出　　处：《自动化与仪器仪表》2025年第2期78-83,共6页Automation & Instrumentation

基　　金：南方电网公司科技项目资助(项目编号:GDKJXM 20230460)。

摘　　要：分布式光伏发电系统的建模对于光伏系统的仿真、评估和控制具有重要意义。与传统的线性系统不同,分布式光伏发电系统具有非线性和多参数的特点,传统的数学模型无法准确识别分布式光伏发电系统。哈默斯坦(Hammerstein)模型是一种典型的块结构非线性模型,广泛应用于各种工业系统的建模中。为了准确、可靠地辨识Hammerstein模型的参数,本文提出了一种基于可分离-随机信号的优化算法。首先,利用实际数据训练分布式光伏发电系统的标称模型,此基础上,基于可分离信号实现线性子系统和非线性子系统的分离辨识。综合实验和测试结果分析表明,基于Hammerstein模型的分布式光伏发电系统具有很强的竞争力,特别是在精度和可靠性方面。The modeling of distributed photovoltaic generation systems is important for the simulation,evaluation and control of photovoltaic systems.Different from the traditional linear system,the distributed photovoltaic generation system has the characteristics of nonlinearity and multi-parameter,and the traditional mathematical model cannot accurately identify the distributed photovoltaic generation system.Hammerstein model is a typical block structure nonlinear model,which is widely used in the modeling of various industrial systems.In order to identify the parameters of Hammerstein model accurately and reliably,this paper proposes an optimiza-tion algorithm based on separable-random signal.Firstly,the nominal model of the distributed photovoltaic power generation system was trained by using the actual data.On this basis,the separation identification of the linear subsystem and the nonlinear subsystem was realized based on the separable signal.Comprehensive experiments and analysis of test results show that the distributed photovol-taic system based on Hammerstein model is very competitive,especially in terms of accuracy and reliability.

关 键 词：分布式光伏发电系统 HAMMERSTEIN模型 可分离-随机信号 参数辨识 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]