【摘要】 研究了自适应噪声对消技术在供电系统谐波电流检测模块中的应用及其使用的传统LMS算法,分析了传统固定步长算法的不足,并借鉴已有的多种变步长函数,提出了一种结合双曲正割和双曲正切变步长函数的自适应滤波改进算法.通过Matlab软件进行实验,分别模拟了采用传统LMS算法、双曲正割函数算法、双曲正切函数算法以及改进后的变步长算法的电流检测过程,并将检测到的基波电流波形进行比较,发现改进算法结合了所选两种函数收敛速度快、稳态精度好的优点,面对系统发生突变有了良好的响应速度和稳定能力,验证了改进算法的合理性.更多还原

【Abstract】 The application of adaptive noise cancellation technology in the harmonic current detection module of power supply system was introduced, and the traditional LMS(Least Mean Square) algorithm with analyzing the shortcomings of the traditional fixed-step algorithm and drawing on the existing variable step size functions were discussed. An improved adaptive filtering algorithm combining hyperbolic secant and hyperbolic tangent step function was proposed. Experiments were carried out by Matlab, and the current detection process using traditional LMS algorithm, hyperbolic secant function algorithm, hyperbolic tangent function algorithm and the improved variable step size algorithm was simulated, and the detected fundamental wave current waveforms were compared. It is found that the improved algorithm combines the advantages of fast convergence and good steady-state accuracy of the selected two functions. It has good response speed and stability when confronts sudden changes in the system, which verifies the rationality of the improved algorithm.更多还原