基于人工神经网络的12扇区SVPWM算法
Twelve-Sector SVPWM Algorithm Based on Artificial Neutral Network
为提高开关管利用率、驱动效率和系统可靠性，本文针对永磁同步电机空间电压矢量脉宽调制算法进行研究。首先理论分析了12扇区的混合SVPWM算法的导通过程，并结合实际PWM生成特性进行了优化，在此基础上提出了一种基于人工神经网络的计算扇区号和导通时间的方法，有效提高了计算效率，并消除了死区时间带来的电流波形畸变和转矩脉动问题，通过仿真验证了该方法的有效性。
In this paper, the best conduction way for SVPWM of PM synchronous motor will be introduced, for improvement of the switch transistor's availability, drive efficiency and system reliability. The paper first analyzes the conduction of hybrid SVPWM algorithm in 12 sectors and optimizes it with PWM actual formation characteristics, based on which a new calculation method is proposed for sector number and conduction time on account of artificial neutral network(ANN) and in this way, calculation efficiency is greatly improved，current harmonic distortion caused by inserting dead-time and torque pulsation are eliminated. The method's validity and effectiveness has been verified through simulation.
吴迪、张幽彤、杨春晖
电机变压器、变流器、电抗器自动化技术、自动化技术设备
动力机械及工程SVPWM12扇区人工神经网络死区时间
power machinery and engineeringSVPWM12 sectorsArtifitial Neutral NetworkDead time
吴迪,张幽彤,杨春晖.基于人工神经网络的12扇区SVPWM算法[EB/OL].(2015-09-15)[2025-08-05].http://www.paper.edu.cn/releasepaper/content/201509-128.点此复制
评论