改进粒子群算法的SHEPWM开关角计算方法研究  

A Computational Method of SHEPWM switching Angles for Cascaded Inverter by Improving PSO

作　　者：吴铁洲[1] 熊金龙 薜竹山 

机构地区：[1]湖北工业大学太阳能高效利用湖北省协同创新中心,湖北武汉430068 [2]国网湖北省电力公司检修公司,湖北武汉430077

出　　处：《湖北工业大学学报》2016年第1期46-50,共5页Journal of Hubei University of Technology

基　　金：国家自然科学基金(51247004);湖北省教育厅科技项目(D20121403)

摘　　要：针对多电平逆变器消除特定谐波中多变量非线性方程组难于获得精确解的问题,以十一电平逆变器为例,建立特定谐波消除脉宽调制(SHEPWM)目标方程,采用改进粒子群优化算法(PSO)求解目标方程,选取非线性时变权重,以平衡算法局部搜索与全局搜索,变异影响全局最优解的随机数,避免算法后期陷入局部最优解。通过Matlab/Simulink试验对改进的粒子群算法、标准粒子群算法与SPWM方法进行比较分析,并得到M=0.85时,改进粒子群的THD%最大降落幅度与SPWM低8.03%;M=0.095时,改进粒子群与标准粒子群相比,THD%降幅最大为33.1%,验证了改进PSO方法能够精确求解SHEPWM方程组。This paper proposes a method based on improved particle swarm algorithm to get the precise so- lution of the target equation owing to the difficulty in solving multi-level cascaded inverter specific harmon- ics elimination equations, taking eleven-level inverter as an example. It selected the nonlinear time-varying inertia weight to balance between portion and overall situation, and altered the random number to avoid lo- cal optima of late algorithm for global optimal solution. Comparative analysis was then conducted by Mat- lab/Simulink test on improved particle swarm optimization, the standard PSO algorithm and SPWM meth- od. The analysis has the following findings. With M = 0.85, compared with SPWM method, the improved particle swarm landed maximum amplitude of THD% is 8.03% ; with M = 0.095, compared with standard PSO, THD% of the improved PSO decline up to 33.1%, so the study proves that improving PS0 can a- chieve the solution to SHE-PWM equations.

关 键 词：多电平逆变器 SHEPWM 粒子群算法 惯性权重 

分 类 号：TM464[电气工程—电器]