摘要
局部放电(PD)是导致变压器、气体绝缘全封闭组合电器、高压电缆等变电装备绝缘故障的主要原因,也是反映变电装备绝缘状态以及研判绝缘老化程度的主要依据。针对现有的超声局放监测存在的传感器灵敏度低、故障定位精度不高等问题,利用多传感器信号融合,研究了一种线性预测残差分析和压扩算法相结合的高精度PD定位检测方法,对多个传感器检测到超声局放脉冲信息之间的到达时间差做出准确估计。在此基础上,研制了一种智能超声局放传感器系统,通过源端局放超声波信号的自适应滤波和放大,提高了传感器的现场抗干扰能力和检测灵敏度。实验结果表明所提出的PD源定位方法的定位精度在1.2 cm内,验证了所提出的智能超声局放传感器的高定位精度。
Partial discharge(PD)is the primary cause of insulation failure of transformer,gas insulated switchgear(GIS),high voltage cable,and other substation equipment.It is also the principal consideration to assess the insulation state of substation equipment and to study the insulation aging degree.Aiming at the disadvantages of low sensor sensitivity and poor fault location accuracy in the existing ultrasonic PD monitoring,a high-precision PD location method based on the combination of linear prediction residual analysis and companding algorithm is studied by using multi-sensor signal fusion,in where the time difference of arrival(TDOA)between the ultrasonic PD pulses detected by multiple sensors is estimated accurately.On this basis,an intelligent ultrasonic partial discharge sensor system is developed,by developing the adaptive filtering and amplification of the ultrasonic signal at the source end,the on-site anti-interference ability and detection sensitivities of the sensor are improved.The experimental results show that the positioning accuracy of the proposed PD source localization method is within 1.2 cm,which demonstrates the high positioning accuracy of the proposed intelligent ultrasonic PD sensor.
作者
尹康
李丽
钟孔露
陈庆伟
Yin Kang;Li Li;Zhong Konglu;Chen Qingwei(Zhejiang Huayun Power Engineering Design Consulting Co.,Ltd.,Hangzhou 310000,China;Economic and Technical Research Institute of State Grid Shandong Electric Power Company,Ji′nan 250021,China)
出处
《能源与环保》
2021年第10期211-217,共7页
CHINA ENERGY AND ENVIRONMENTAL PROTECTION
基金
国家电网公司总部科技项目(5200-202055134A-0-0-00)。
关键词
高精度定位
局部放电
线性预测残差分析
扩压算法
超声传感
high precision positioning
partial discharge
linear prediction residual analysis
companding algorithm
ultrasonic sensor
