频率变化时交流采样算法分析

Research of AC sampling when frequency is unsteady

电力系统交流采样中使用均方根法计算时必须取得交流信号1个整周期的数据,否则会产生较大的误差。分析了引起误差的原因,使用Matlab的仿真计算,分析了误差大小和采样点数之间的关系。在此基础上提出了2种改进的算法,分别利用有效值作为参考选取起始点和数值补偿进行误差校正。前者在频率变化不大时,无需准确知道系统频率就能保持较高计算精度,但在采样频率较低时误差较大;后者计算精度高,但必须准确知道频率值。

Mean square root method is used in AC sampling computing of power system, in whichthe data points covering exactly the whole cycle are needed,otherwise considerable errors will occur when system frequency is changing. The factors that cause the errors are analyzed and the relationships between errors and sample points are simulated with Matlab,based on which two improved computing methods are proposed to increase the precision,they are the start point selection according to effective value and the error correction with numerical compensation. The former does not need the precise system frequency and has higher calculation accuracy when system frequency changes little. But considerable error will occur when the sampling frequency is low. The latter has even higher accuracy,but the system frequency must be known.