一种采用虚拟磁链模型预测的新型PWM算法  

A Novel Algorithm of PWM Using Virtual Flux Model Prediction

作　　者：杜少通[1] 伍小杰[1] 周娟[1] 王福均[1] 杨宇[1] 

机构地区：[1]中国矿业大学信息与电气工程学院,江苏省徐州市221116

出　　处：《中国电机工程学报》2015年第3期688-694,共7页Proceedings of the CSEE

基　　金：国家自然科学基金项目(51377160)~~

摘　　要：针对变流器脉宽调制(pulse-width modulation,PWM),基于虚拟磁链的模型预测,提出一种新型脉宽调制算法(novel algorithm of pulse-width modulation,NPWM)。建立由正弦调制电压和变流器输出电压构成的虚拟磁链模型,以磁链误差最小为优化目标建立价值函数,采用模型预测获取最优目标电压矢量。基于单一目标预测方法,提出一种多目标预测方案,确保在磁链误差范围一定的前提下降低器件平均开关频率,依据建立的价值函数分析算法流程。仿真和实验比较两种预测算法下NPWM调制和空间矢量脉宽调制的控制效果,验证两种预测方法下NPWM算法的有效性。In allusion to pulse-width modulation（PWM）, a novel algorithm of pulse width modulation（NPWM） was proposed based on virtual flux model prediction. The virtual flux models consisting of sinusoidal modulation voltage and converter output voltage were established. The cost function was built aiming at making the flux error minimum. And the optimal objective voltage vector was obtained by using model prediction. On the basis of single target predictive method, a type of multi-target predictive method was put forward whi＇ch ensures that the device＇s average switching frequency was reduced under the premise of certain range of flux errors. The algorithm flow was analyzed according to the established cost function. The control effect of NPWM and space vector pulse width modulation was compared by simulations and experiments, which verifies the effectiveness of NPWM in the two predictive methods.

关 键 词：脉宽调制 模型预测 虚拟磁链 单目标优化 多目标优化 

分 类 号：TM46[电气工程—电器]