一种改善农电网电能质量的谐波检测算法

An Harmonic Detection Algorithm for Improving the Power Quality in Rural Power Grid

为了提高农村配电网的电能质量和抑制谐波污染,提出了一种应用于APF谐波检测的新的变步长自适应滤波算法。该算法通过当前误差变化值同稳态临界阈值相比较来确定滤波器权系数的更新方式。当误差变化超过稳态临界阈值时,使用步长因子与Sigmoid函数的关系来更新步长,以此来提高算法的初始收敛速度和动态响应速度;相反,步长因子由当前误差与上一步误差的自相关估计来决定,使步长因子只取决于真正的系统跟踪误差,提高了稳态精度。实验表明,该算法能够同时保证谐波检测较快的收敛速度、较高的稳态精度和动态跟踪效果。

In order to improve the quality of rural power distribution network and inhibit the harmonic pollution,a novel variable step size adaptive filtering algorithm is proposed to apply to APF harmonic detection.The algorithm determines the update method of the filter weight coefficient by comparing the variation value of current error with the steady-state critical thresholds.When the error variation beyond the steady-state critical thresholds,the step size is updated by using the relation between step factor and Sigmoid function,so the algorithm is improved in the initial convergence and the dynamic response speed;On the contrary,the step size is determined by the autocorrelation estimation between the current error and the previous error.So the step factor is only determined by the real system tracking error,and the steady-state precision is improved.The experimental results indicated that this algorithm ensure the convergence rate,steady-state accuracy and dynamic tracking performance.