电晕放电会对高压输电线路造成极大危害,因此,检测电晕故障对于电力系统的安全具有重大意义.为了准确定位高压设备电晕放电的故障点,提出了一种基于IHS(Intensity,Hue,Saturation)和小波变换的可见光与紫外光的图像融合算法,首先对可见光进行IHS变换,将得到的I分量与紫外光进行小波分解得到各自的高低频分量,对于低频分量采取加权融合的融合算法,高频分量则采取基于区域特性的融合算法,然后通过小波重构得到新的高低频分量,最后进行IHS逆变换产生新的融合图像.实验结果表明,文中方法在可见光与紫外光融合处理中取得了较好的融合效果,优于IHS变换算法与传统的小波变换算法,且图像更加清晰,细节纹理也更加丰富,有效地保留了边缘信息.

电力系统中局放信号的采集容易出现噪声干扰,从而导致无法准确提取局放信号,针对这一问题提出了联合变分模态分解(VMD)与改进小波阈值的去噪方法.首先将染噪的局放信号进行VMD分解,通过局放信号的重构从而滤除周期性的窄带干扰信号;在VMD的分解以及信号的重构中借助瞬时频率均值法确定K个模态分量,依据峭度准则进行局放信号的重构.对于还存有的部分高斯白噪声提出一种新的小波阈值函数,通过改进的小波阈值法进行高斯噪声的滤除.通过对局放信号处理的仿真,结果显示该方法能够较好地滤除白噪声和周期窄带干扰噪声.在实际的局放实验中,也证实该方法对于局放信号的去噪效果显著,明显提高了信号的信噪比,且比较完整地保留了有效信号.

针对电力电缆中间接头局部放电信息检测系统实际采集到的局部放电信号含有噪声的问题,提出了一种将快速傅里叶变换与改进小渡包变换相结合的处理方法,对于周期性窄带干扰,选取快速傅里叶变换来处理;对于白噪声,通过一种改进的阈值函数的小波包算法进行处理.实际应用结果表明,该方法去噪效果明显,不仅有效去除了局部放电脉冲信号中的噪声,可较好地保留了原始信号的有用信息.

Because of the low current value of series arc fault in AC low voltage circuits and the normal current of some power electronic loads is similar to arc fault current, it is difficult to diagnose the series arc fault accurately. This paper presents a method of series arc fault analysis and diagnosis based on the wavelet transform and difference-energy. Mallat method is used to decompose the signals which are denoised by the wavelet threshold principle. Extract the layers in multi-resolution analysis results which contain much arc fault information and are not affected by the loads frequency for wavelet reconstruction. Then we apply the difference-energy method to diagnosis the signals after reconstruction. The accuracy of the algorithm is verified by the self-made experimental equipment.