有源电力滤波器自适应谐波检测新算法

New Adaptive Harmonic Detection Algorithm of Active Power Filter

为了克服定步长算法无法兼顾快速性和稳态精度的局限,改进现有的变步长自适应算法中存在计算量大、未知参数多的缺点,该文基于均方误差理论推导了一种新的根据误差实时自动调节步长的迭代算法,并应用于有源电力滤波器谐波检测。算法优点在于计算量小,从均方误差最快下降梯度入手,在误差较大时采用大步长,保证快速性;在误差较小时采用小步长,保证稳态精度。Matlab仿真和系统实验验证了文中算法的快速性和准确性,以及应用于有源电力滤波器谐波检测时的有效性。

Since the fixed-step algorithm has a limitation that it cannot keep a balance between steady accuracy and rapidity, and the existing variable-step algorithm has the disadvantages of heavy computation load and numerous unknown parameters, a new automatic adjustment iterative algorithm according to the error is deduced based on the theory of mean square error（MSE）, and it is applied to the harmonic detection of active power filter（APF）. The proposed algorithm has the advantage of less computation load, and based on the fastest decline of MSE, it adopts a large step when the error is bigger to ensure its rapidity, and uses a small step when the error is smaller to guarantee the steady accuracy. The rapidity and accuracy of the proposed algorithm, and the feasibility of its application to the harmonic detection of APF, are verified by Matlab simulation and experiment.