基于伪线性回归算法的自适应滤波算法

为抑制谐振接地系统对地电容中的工频信号和其它谐波成分，提出了一种基于伪线性回归算法的自适应滤波算法。该算法以经平滑滤波后的信号输出误差为代价函数调整滤波器系数，当滤波器输入信号变化时可保持最佳输出信号，在一定程度上补偿了滤波器元件参数变化带来的运算误差。仿真结果表明，与其它自适应算法相比，该算法收敛速度快且失调率低，基于该算法的滤波器结构简单、测量误差较低、可靠性较高。

An Adaptive Filtering Algorithm Based on Pseudo Linear Regression Algorithm

To suppress the power frequency signal and other harmonic components in earth capacitance of resonant grounded system, an adaptive filtering algorithm based on pseudo linear regression algorithm is proposed. Taking the output error of smoothly filtered signal as cost function in the proposed algorithm, the filter coefficient is adjusted; thus the optimal output signal can be kept while input signal of filter changes; the arithmetic error connected with the parameter variation of filter element can be compensated in a certain extent. Simulation results show that comparing with other adaptive algorithms, the proposed algorithm can converge rapidly and its misadjustment rate is low. The filter based on the proposed algorithm possesses following advantages: simple structure, lower measuring error and higher reliability, and it can be used in signal injection.