并网变流器全工况阻抗模型的黑箱辨识方法OA北大核心CSTPCD
Black-box identification of a full operating condition impedance model of grid-tied converter
阻抗分析法因适用于并网变流器控制结构和参数均未知的黑箱系统,已成为工程中评估宽频振荡风险的重要选择.以跟网型三相两电平变流器为例,首先,使用谐波线性化方法,严格推导了计及直流侧耦合的交流侧全工况开/闭环理论阻抗模型,实现了模型参数与工作点高阶项的完全解耦.其次,依据理论模型特征,明确了黑箱辨识方法的基本原理,分析了辨识单一开/闭环阻抗模型频率响应所需的工况数量.使用辨识得到的全工况阻抗模型,实现了不同短路比条件下系统安全运行域的快速评估,并通过时域仿真验证了其准确性.此外,讨论了所提方法的合理性、普适性以及相对于深度学习方法的优越性,为阻抗分析法的实际应用提供了参考.
Impedance-based analysis has become an important choice for engineering for evaluating the risk of broadband oscillation.This is because it adapts to the black-box system with unknown parameters or control structure of grid-tied converters.In this paper,the grid-following three-phase two-level converter is taken as an example.First,using harmonic linearization,the open/closed-loop theoretical full operating condition AC impedance model(IM)considering the DC coupling is rigorously deduced.In particular,the model parameters and the high-order terms of operating points are fully decoupled.Secondly,based on the features of the theoretical model,the basic principle of the black box identification method is clarified,and the numbers of operating conditions for identifying a single frequency response of the open/closed-loop IM is analyzed.Using the identified full operating condition IMs,rapid evaluation of the safe operation area of the system under various short-circuit ratios is realized,and the accuracy is verified by time domain simulations.In addition,the rationality and universality of the proposed method as well as its superiority over deep learning methods are discussed.This provides a reference for the practical application of the impedance-based method.
赵崇滨;姜齐荣
电力系统及发电设备控制与仿真国家重点实验室(清华大学电机工程与应用电子技术系),北京 100084
阻抗分析法黑箱系统模型辨识全工况稳定性分析
impedance-based analysisblack-box systemmodel identificationfull operating conditionstability analysis
《电力系统保护与控制》 2024 (004)
48-55 / 8
This work is supported by the National Natural Science Foundation of China and Smart Grid Joint Fund Integration Project(No.U22B6008). 国家自然科学基金智能电网联合基金集成项目资助(U22B6008)
评论