基于高低频特征融合的电力设备热故障模糊检测方法

Fuzzy Detection Method for Thermal Faults of Power Equipment Based on High and Low Frequency Feature Fusion

电力设备在运行过程中,受电流和电压的影响易出现热故障,使成像特征为高低频混合状态,现有方法仅检测高频特征或低频特征,导致检测结果置信度低、检测效果差。提出了基于高低频特征融合的电力设备热故障模糊检测方法。采用红外成像传感器采集电力设备图像,利用区域能量自适应加权法,处理图像低频和高频特征。通过小波逆变换融合高低频特征计算小波系数。利用灰度值提取特征参数,得到热故障检测结果。实验结果表明:所提方法可精准检测热故障区域,置信度最高值为0.96,检测准确率高达90%,平均交并比为89%,具有较好的热故障检测效果。

During the operation of power equipment,thermal faults are easy to occur due to the influence of current and voltage,which makes the imaging feature a mixed state of high frequency and low frequency.Existing methods only detect high-frequency features or low-frequency features,resulting in low confidence of detection results and poor detection effect.A fuzzy detection method for thermal fault of power equipment based on high and low frequency feature fusion is proposed.The infrared imaging sensor is used to collect the image of power equipment,and the regional energy adaptive weighting method is used to process the low frequency and high frequency features of the image.The wavelet coefficients are calculated by fusing high and low frequency features with inverse wavelet transform.The feature parameters are extracted by gray value,and the thermal fault detection results are obtained.The experimental results show that the proposed method can accurately detect the hot fault area,the highest confidence level is 0.96,the detection accuracy is as high as 90%,and the average crossover/merge ratio is 89%,which has a good hot fault detection effect.