基于递推最小二乘的光伏逆变器动态模型辨识

DYNAMIC MODEL IDENTIFICATION OF PHOTOVOLTAIC INVERTER BASED ON RECURSIVE LEAST SQUARES

提出一种基于递推最小二乘的光伏逆变器动态模型辨识方法,不考虑三相光伏逆变器及其控制系统的拓扑结构,将其整体视为一个"黑箱",通过对输入量、输出量的分析将其等效为双输入多输出系统,利用递推最小二乘法在线实时辨识光伏逆变器的动态模型及参数。针对夏季、冬季不同典型天气下的数据集建立光伏逆变器等效动态模型,模型输出与测试数据误差较小,验证辨识方法的精确性与适用性。另外,模型最优阶较低这一特性也降低内存占用量,提高辨识速度。

This paper proposes a dynamic model identification method for the PV inverter based on recursive least squares(RLS). Without consideration of the topology of three-phase PV inverter and its control system,the whole inverter section is regarded as a"black box". Through the analysis of input and output,it can be equivalent to a doubleinput-multi-output system. RLS method is used to identify the PV inverter dynamic model and parameters. According to the data sets of different typical weathers in summer and winter,equivalent dynamic models of PV inverter are established respectively. The error between the model output and the test data is small,which verifies the accuracy and applicability of the identification method. Moreover,the lower order of the model reduces the memory footprint and improves the identification speed.