【摘要】 设计了一种基于系数可调的有限冲激响应(FIR)滤波器和最小均方误差(LMS)算法的自适应数字滤波器.该自适应滤波器具有运算速度快、计算量小、稳定性好、不需要关于输入信号的先验知识等优点,非常适合于实时滤波.设计了自适应滤波器的程序流程和验证滤波器效果的实验方法,并通过模拟实验确定了适合于在线监测的自适应参数和FIR滤波器阶数.通过现场实测,证明了滤波器的有效性.更多还原

【Abstract】 An adaptive filter is designed and employed to extract PD signal submerged in wideband interference sinusoidal signals. The adaptive filter, based on FIR filter and leastmeansquare (LMS) error algorithm, has some advantages, such as less calculation time, minimum memory, better stability, and ignoring prior knowledge of input signal. The program flow of filter and simulation method used to verify the filtering effects are designed. Moreover, the adaptive parameter μ and number of FIR filter weights N, which meet the requirements of online PD monitoring, are also decided by simulation of computer. The results measured from laboratory and field test show that the adaptive filter is effective and especially satisfies online PD monitoring.更多还原