红外图像增强机车电力系统故障诊断与定位

Fault diagnosis and location for locomotive electric power system based on infrared image enhancement

针对机车电力设备红外图像的对比度较差、细节不够明显的特点,该文提出一种基于非线性变换(nonsubsampled contourlet transform,NSCT)的图像增强算法,在算法模型中构造一个非线性匹配函数,该函数能够对红外图像的边缘进行一定的增强,并且能够抑制噪声的干扰。实验结果表明:该方法能够对电力系统过热故障进行智能诊断,并且能够提供复选解决方案。

For locomotive power equipment infrared image contrast is poorer, and its detail characteristics are not obvious enough, this paper puts forward a kind of the image enhancement algorithm based on NSCT(nonsubsampled contourlet transform) nonlinear transform. The algorithm constructs a nonlinear matching function which can enhance the edge of the infrared image and suppress the interference of noise. Experimental results show that the proposed method can overheat faults in the power system with intelligent diagnosis, and this study can provide a solution for checking.