Robust Model Predictive Current Control of Three-Phase Voltage Source PWM Rectifier With Online Disturbance Observation
Robust Model Predictive Current Control of Three-Phase Voltage Source PWM Rectifier With Online Disturbance Observation
复制标题
DOI：
10.1109/tii.2012.2187912
复制
发表时间：
2012-03
影响因子：
12.3
通讯作者：
C. Xia;Meng Wang;Zhanfeng Song;Tao Liu
中科院分区：
文献类型：
--
作者：
C. Xia;Meng Wang;Zhanfeng Song;Tao Liu
针对三相电压型PWM整流器，提出了一种带干扰观测器的鲁棒模型预测电流控制器。新算法以恒定开关频率(CF-MPC)运行。为了在每个采样周期内最小化d轴和q轴的瞬时电流误差，通过选择合适的电压矢量序列和计算占空比来实现CF-MPC。讨论了该算法的基本原理，推导了在6个不同扇区施加每个变流器电压矢量时，d轴和q轴电流的瞬时变化率。针对影响预测控制性能的参数失配和模型不确定性，构造了一种Luenberger观测器。扰动观测器的增益由根轨迹分析确定。此外，还分析了当电感滤波参数存在误差时干扰观测器的稳定性。由于采用了MPC控制器，该方法具有固有的快速动态响应，同时由于采用了组合观测算法，对扰动具有较强的控制性能。在一台1.1kW的VSR上进行了仿真和实验，验证了该方法的有效性。
This paper presents a robust model predictive current controller with a disturbance observer (DO-MPC) for three-phase voltage source PWM rectifier. The new algorithm is operated with constant switching frequency (CF-MPC). In order to minimize instantaneous d- and q-axes current errors in every sampling period, CF-MPC is implemented by selecting appropriate voltage vector sequence and calculating duty cycles. The fundamental of this algorithm is discussed and the instantaneous variation rates of d- and q-axes currents are deduced when each converter voltage vector is applied in six different sectors. A Luenberger observer is constructed for parameter mismatch and model uncertainty which affect the performance of the MPC. The gains of the disturbance observer are determined by root-locus analysis. Moreover, the stability of the disturbance observer is analyzed when there are errors in the inductor filter parameter. The proposed method has an inherent rapid dynamic response as a result of the MPC controller, as well as robust control performance with respect to the disturbance due to use of the combined observation algorithm. Simulation and experimental results on a 1.1 kW VSR are conducted to validate the effectiveness of the proposed solution.