基于反向传播神经网络的SVM技术在电压型变流器中的应用研究

在分析三相电压型变流器空间矢量调制(SVM)技术基本原理的基础上，提出了一种基于反向传播神经网络结构的SVM技术(CPN—SVM)的实现方法。该方法采用CPN竞争层来计算SVM中各个矢量的具体作用时间，避免了计算正弦函数这一非线性运算，缩短了计算时间，采样周期的可进一步缩短。仿真和实验表明：CPN—SVM除了具备SVM的固有优点外，还有下述几个显著优点：①大大降低了整个控制系统的软硬件成本，提高了对开关瞬态位置判断的准确性；②随着采用周期的缩短，最大开关频率相应增大，从而提高了整个系统的传输带宽③避免了由于计算误差给SVM波形中所带来的附带谐波。

RESEARCH ON CPN-SVM TECHNIQUE APPLIED IN VSC

Based on the principle of SVM for three-phase Voltage Source Converter, this paper presents a Counter Propagation Neural Networks SVM implementation on basis of a neural network structure. In order to avoid the sin function, the proposed technique takes advantage of a competitive layer to calculate the duration of the adjacent switching state vectors. This technique reduces hardware and software complexity, and computation time, and improves the accuracy of the positioning of the switching instants. The technique exhibits the following features: a) possibility of higher switching frequencies, b) higher bandwidth of the control loops, c) hardware and software saving, d) reduction of parasitic harmonics in all PWM waveforms. All this had been verified by simulation and experimental results.