摘要
电力电子变换器是频域内典型的单输入多输出系统,当输入某一频率的扰动信号时,变换器各状态变量既包含扰动频率分量,也包含与扰动相关的边带频率成分。在包含多个电力电子变换器的分布式供电系统中,一个变换器的开关纹波为另一个变换器的扰动,这种相互作用在某些情况下可能会导致母线电压差频振荡从而影响系统电能质量。然而,传统小信号模型以单个变换器的分析和设计为背景提出,主要用于描述变换器的低频特性。由于这些模型忽略了开关变换器的很多固有特性,因此不能准确地分析上述变换器之间在开关频率附近的相互作用。为此,提出了一种新的矩阵小信号模型并以Buck变换器为例进行了详细的说明。该模型可以准确地描述变换器的单输入多输出特性,并解释分布式供电系统中电力电子变换器相互作用导致的母线电压差频振荡现象。对比结果表明,传统的平均小信号模型和多频率小信号模型都是所提出的矩阵模型在不同情况下的近似。仿真和实验结果证明了所提模型的准确性和有效性。
Power electronic converters are typical single input multiple output (SIMO) systems in frequency domain. When a perturbation frequency is input, the state variables contain not only the perturbation frequency but also its sidebands. In distributed power system (DPS), the switching frequency ripples of one converter are the perturbations of the other converter. Such interaction may introduce thebeat frequency oscillation. However, traditional small-signal models mainly focus on the design and analysis of single converter, which ignore some inherent features of the switching converters. These models are questionable to analyze the interactions of power converter around their switching frequencies. Therefore, a matrix-basedmulti-frequency model is proposed. The voltage mode controlled buck converters are taken as examples. The proposedmodel could describe the SIMO characteristics of power electronic converters and explain the beat frequencyoscillations in DPS, while the traditional models fail to do. Detailed comparison shows that traditionalsmall-signal modelsare thesimplifications of the proposed model in different situations. The simulation and experimental results validate the proposed model.
作者
岳小龙
卓放
杨书豪
裴云庆
孙力
孙亮
Yue Xiaolong;Zhuo Fang;Yang Shuhao;Pei Yunqing;Sun Li;Sun Liang(State Key Laboratory of Electrical Insulation and Power Equipment Xi’an Jiaotong University Xi’an 710049 China)
出处
《电工技术学报》
EI
CSCD
北大核心
2017年第4期250-259,共10页
Transactions of China Electrotechnical Society
基金
国家高科技研究发展计划(863计划)(2015AA050606)
国家自然科学基金(51177130)资助项目
关键词
分布式供电系统
BUCK变换器
多频率矩阵模型
差频振荡
Distributed power system
Buck converter
multifrequency matrix model
beatfrequency oscillation