【摘要】 GIL是一种快速发展的新型输电设备,但目前其故障的检测是一个难点。根据现场运行情况统计,运行中的GIL设备故障主要分为机械性故障和放电性故障2类,其中绝缘子的损坏和金属焊点松动是机械故障常见问题,反映在振动谱图上主要集中于低频部分。基于振动检测的原理,研究低频振动信号的特征参数提取,阐述样本处理和特征参数归一化的具体实现步骤,建立基于SVM的核算法诊断模型,为GIL机械类故障的诊断提供理论支撑。更多还原

【Abstract】 GIL is new a fast-developing transmission device and its faults detection is still a difficult problem. From the statistics of field operation, it is found that the most common running faults are mechanical faults and discharge faults. This paper focuses on the fault detection of mechanical faults, which might be caused by the insulator damage and loosening of metal solder joint. This type of fault is reflected in the low frequency part of vibration spectrum. Under this background, the extraction of characteristic parameters of low frequency vibration signal is studied concerning the principle of vibration monitoring. Then, the procedure of sample processing and data normalization of vibration signal is introduced in detail. A fault detection model is successfully constructed based on a model of SVM fault accounting. Finally, the model is verified by the comparison of parameter samples and a calculation group. This research provides theoretical support for GIL mechanical fault diagnosis.更多还原