局部放电在线监测中的自适应滤波方法

ADAPTIVE FILTER METHOD FOR PARTIAL DISCHARGE ON-LINE MEASUREMENT SYSTEM

局部放电在线监测对保证电力变压器安全运行有重要意义，实现在线监测的关键是从强干扰中监测出微弱的放电信号。针对怎样抑制周期性干扰，文献中报道较多的是LMS自适应滤波方法，但该方法需调整参数多，对脉冲型干扰表现出不稳定性。为此，文中分别论述并比较了LMS算法和基于鲁棒RLS算法的自适应滤波方法的基本原理及实现方法。实验结果表明：鲁棒RLS算法较好地解决了LMS算法存在的问题，且能获得更高的信噪比，基本适用于局部放电在线监测。

On-line measurement of partial discharge (PD) is important to ensure safe operation for power transformer. The key of on-line measurement is to pick up discharge signal under strong interference. Aiming at how to prevent periodic interference, more papers have covered LMS adaptive filter, which needs to adjust many parameters and shows instability in pulse type interference. This paper discusses the fundamental principles and the design methods of LMS algorithm and robust RLS algorithm. The experiment results show that robust RLS algorithm can solve the problem existing in LMS algorithm and gain higher signal-to-noise ratio. Therefore. it can be applied in PD on-line measurement.