基于双矢量模型预测直接功率控制的双馈电机并网及发电
发布时间:2018-03-30 09:24
本文选题:双馈电机 切入点:并网同步 出处:《电工技术学报》2016年05期
【摘要】:模型预测直接功率控制(MPDPC)以其原理简单、动态响应快和控制目标灵活等优点在并网变换器控制中得到了广泛关注。它利用系统模型来预测下一时刻的有功功率和无功功率,通过枚举比较得到使给定功率和反馈功率之间误差最小的电压矢量,相比传统的基于矢量表的直接功率控制具有更好的动态和稳态性能。为了进一步提高稳态性能并降低采样频率,提出一种基于双矢量的改进模型预测直接功率控制,即把一个控制周期分配给一个非零矢量和一个零矢量,通过优化非零矢量的作用时间来进一步减小功率脉动。通过合理安排矢量作用顺序甚至可以获得比传统MPDPC更低的开关频率而且稳态性能更好。在15 k W双馈风力发电系统上的仿真结果表明,相比传统MPDPC,所提出的双矢量MPDPC能够以更低的采样频率实现快速平滑的并网同步和灵活的有功、无功控制,并且开关频率更低,具有较大的实用价值。
[Abstract]:Model Predictive Direct Power Control (MPDPC) is widely used in grid-connected converter control because of its simple principle, fast dynamic response and flexible control target. It uses system model to predict active and reactive power at the next moment. The voltage vector with minimum error between the given power and the feedback power is obtained by enumeration and comparison. Compared with the traditional direct power control based on vector table, it has better dynamic and steady-state performance. In order to further improve the steady-state performance and reduce the sampling frequency, an improved model based on double vectors is proposed to predict the direct power control. That is, to assign a control period to a nonzero vector and a zero vector, By optimizing the action time of nonzero vectors, the power pulsation can be further reduced. Even lower switching frequency and better steady-state performance than conventional MPDPC can be obtained by properly arranging the sequence of action of vectors. At 15kW doubly-fed wind power, the power pulsation can be further reduced. The simulation results on the power generation system show that, Compared with the traditional MPDPCs, the proposed dual-vector MPDPC can realize fast and smooth grid-connected synchronization and flexible active and reactive power control with lower sampling frequency, and the switching frequency is lower, which is of great practical value.
【作者单位】: 北方工业大学电力电子与电气传动北京市工程研究中心;
【基金】:国家自然科学基金(51207003、51577003) 北京市科技新星计划(xx2013001)资助项目
【分类号】:TM614
,
本文编号:1685312
本文链接:https://www.wllwen.com/kejilunwen/dianlidianqilunwen/1685312.html