一种空间矢量的三相逆变器调制方法研究

A Three-Phase Inverter Strategy Research with Space Vector Modulation

针对三相逆变器提出了一种基于PID神经网络的空间矢量调制(SVPWM)方法。三相逆变器具有矢量的开关状态,其调制算法的实现非常复杂,在设计的方案中,摒弃传统的神经网络,采用PID神经网络用于SVPWM算法的快速实现,三层前馈PID神经网络以指令电压和角度作为输入信息,借助于一个简单的计时器和逻辑电路,产生三相12路对称脉宽调制波形。最后利用Matlab神经网络信息工具箱依据误差逆变算法使用实测数据对该系统进行了测试,结果证明,在全部线性调制区内,系统性能很好。

As for the three-phase inverter, this paper introduces a space vector pulse width modulation （SVPWM） based on PID neural network. A three-phase has the switch state of vector and the implementation of its modulation algorithm is very complex. This research adopts PID neural network modularization design for rapid implementation of SVPWM algorithm instead of the traditional neural network. The three-layer feed forward PID neural network, produces three-phase symmetrical 12 sinusoidal pulse width modulation waveforms with the instruction voltage and angle as the input information and with the aid of a simple timer and logic circuit. The measured data are used to test the proposed system by a back propagation algorithm in the MATLAB Neural Network Toolbox. Experimental results show that the method is feasible, and can effectively improve response speed and stability in all linear regions.