基于在线预测的新型dq0变换DVR检测算法

Detection algorithm with dq0 transform based on on -line prediction for dynamic voltage restorer

针对目前常用的动态电压恢复器(DVR)检测算法难以满足DVR实时性的要求,提出一种基于在线预测的新型dq0变换检测方法。利用在线检测所得的当前时刻前数个采样点信息来还原当前时刻的基波信息,有效避免了通常dq0算法中延时较大的数字滤波器环节。将谐波补偿量与基波补偿量相加之和作为DVR输出的补偿参考波信号,保证了负荷侧电压保持稳定。仿真分析验证了该方法的正确性,实验结果表明该算法可在1ms内检测到电压扰动,并输出补偿参考电压信号,整个DVR系统能在3ms内输出补偿电压信号,满足了DVR实时补偿的需要。与采用数学变换结合低通滤波检测算法相比,该算法具有实现容易、运算速度快等优点。

In order to meet the real time requirement of DVR（Dynamic Voltage Restorer）,a voltage sag detection algorithm with dq0 transform based on on- line prediction is presented. The present fundamental wave is predicted with a few values sampled just before this moment,the filters with big time- lag in general dq0 algorithm are thus avoided. The sum of harmonic compensation and fundamental compensation is taken as the reference wave to keep the load voltage stable. Simulative analysis shows its validity and experimental results indicate that,it ,detects voltage sags within 1 ms and DVR system outputs compensation voltage within 3 ms,which meets the real time requirement of DVR. Compared with the detection algorithms integrating mathematic transform and lowpasses,this algorithm is easy in implementation and fast in calculation.