摘要
电压闪变是严重的电能质量问题之一,本文提出采用基于可变遗忘因子的高阶累积量递推最小二乘(RLS)算法跟踪电压闪变包络。基于高阶累积量的误差准则取代传统的基于二阶统计量的误差准则,使算法不受任何高斯噪声的影响;可变遗忘因子的应用,使得算法能够快速跟踪包络的变化,又能在稳态情况下具有较好的收敛性能。Matlab仿真分析结果表明,在高斯噪声污染存在的情况下,本文算法相比传统RLS算法具有更高的估计精度,能够在噪声中准确地检测出电压闪变包络。
Voltage flicker is a serious problem for electric power quality,and the higher-order cumulant RLS algorithm based on variable forgetting factor is introduced in this paper as effective approach for tracking the voltage flicker envelope.With the improved RLS algorithm,the error criterion based on conventional mean-squared-error(MSE)is replaced by the higher-order cumulant MSE criterion,which makes the algorithm not influence by the Gaussian noise. The application of variable forgetting factor(VFF)not only can improve the speed of algorithm to track the variation of envelope,but also assure better convergence performance under steady state.Simulations by Matlab shows that the proposed algorithm has better estimation accuracy compared with that of the conventional RLS algorithm under the disturbing of gaussian noise,and can detect voltage flicker envelope accurately.
出处
《现代电力》
2011年第5期24-28,共5页
Modern Electric Power
关键词
高斯噪声
高阶累积量
RLS算法
可变遗忘因子
电压闪变包络
Gaussian noise
higher-order cumulant
RLS algorithm
variable forgetting factor
voltage flicker envelope