基于AFE变换器的改进模型预测控制策略

Improved Model Predictive Control Strategy Based on AFE Converter

针对有源前端AFE(active front-end)变换器电流环采用的有限控制集模型预测控制算法FCS-MPC(finite control set model predictive control),存在数字控制器计算量大、输出的电流脉动大和系统性能差等问题,提出了一种改进模型预测控制策略。首先通过坐标变换构建α-β坐标下AFE变换器的预测模型;结合电压型空间矢量变换的原理,引入矢量扇区判断法,减少最优电压矢量选择的次数;最后根据选出来的电压矢量,在1个控制周期内同时施加有效电压矢量和零电压矢量,并分配相应的作用时间。该方法不但保留了传统MPC算法的快速动态响应特性,而且改善了系统的稳态性能。仿真和实验结果均验证了所提方法的正确性与有效性。

When the finite control set model predictive control(FCS-MPC)algorithm is applied to the current loop in an active front-end(AFE)converter,there will exist problems such as large computational complexity of the digital contro-ller,large output current ripple,and poor system performance.To solve these problems,an improved MPC strategy is proposed.First,the prediction model of the AFE converter under theα-βcoordinate is constructed through coordinate transformation.Then,combined with the principle of voltage space vector equivalent transformation,the sector discrimination method based on vectors is introduced to reduce the number of selections of the optimal voltage vector.Finally,according to the selected voltage vector,the effective voltage vector and zero-voltage vector are simultaneously applied in a control cycle,and the corresponding acting time is assigned.This method not only retains the fast dynamic response characteristics of the traditional MPC algorithm,but also improves the system’s steady-state performance.Both simulation and experimental results verified the correctness and validity of the proposed method.