摘要
分析了空间矢量脉宽调制(SVPWM)的基本原理,在此基础上提出了一种基于复合人工神经网络(ANN)的SVPWM新算法,充分地利用ANN的快速并行处理能力、学习能力,缩短了计算时间,减低了由控制延时引起的谐波成分。最后在MATLAB/Simulink环境下结合ANN工具箱建立了仿真模型,在感应电机直接磁场定向的矢量控制系统中进行了仿真研究,仿真结果证明了基于复合网络的SVPWM新算法的可行性。
The basic principle of space vector pulse-width-modulation （SVPWM） was analyzed, and a novel algorithm of SVPWM based on the multiple artificial-neural-network （ANN） was proposed. The calculation time is shortened, and the harmonic due to control delay is reduced because of the fast parallel computation and learning capability of ANN. A computer simulation model for direct vector control system of induction motor was developed using MATLAB/Simulink together with the Neural Network Toolbox. The simulation results demonstrate the feasibility of Multiple-ANN-SVPWM.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2006年第2期420-423,共4页
Journal of System Simulation
基金
教育部科学技术研究项目(03085)
