Adaptive Filter In PD Measurement And Experiment Research

In recent years, with the improvement of capacity and voltage-level of HV equipments, it becomes more and more important to keep HV equipment safely running in employed years. In addition, insulation aging takes most of proportion of HV equipment accident. Therefore, to keep HV equipment in good condition, firstly, it should keep insulation in good condition. Then, we need a way that can measure HV equipment's insulation effectively, credibly and conveniently. As an important way to measure insulation condition, PD measurements take more and more attention from researchers, and now, it takes great development both in detecting theory and detecting technology. It is also employed by factory and power station for some useful measurements way. But, there are still some problems in PD measurement, the first is noise disturbing and the second is the true relation between PD mechanism and insulation aging. This paper is aim to solve above problems. With the help of computer, PD filtering problems are researched, and in addition, it is also discussed the relationship between PD spectrum and insulation aging and defects by way of PD experiment. Firstly, this paper designs a LMS algorithm adaptive filter to solve period noise disturbing. Compared with normal noise restraining, adaptive filter have many advantages: it need not know knowledge of input signal, and it can also change itself to meet changing of input signal. According to computer simulation, the way, which employed by this paper, is simple, stable and computing fast. In condition of true selected filter coefficients, it can usefully solve period noise disturbing in PD measurement. Secondly, PD experiments are carried in perfect LDPE sample, artificial gaps LDPE sample, and water-tree LDPE. According to PD spectrum in different samples and different test conditions, the relation between the characteristics of PD spectrum and insulation inner defects are discussed.