Transformer Insulation Monitoring System For The Research And Design

With the rapid development of electrical engineering, the demand on the service reliability of power system has become higher and higher. Electric transformers play very important roles in power system, and the dielectric loss tangent is an important parameter to reflect the insulation condition ,but nowadays, traditional preventing maintenance can't follow the rapid development of the power system. The on line method not only can ensure the reliability by finding out the latent troubles and preventing accidents, but also reduce the loss of the power cut and save a lot of budget because we don't need to check the system frequently.Direct measurement of loss is Zero cross point comparative method. Another one is to adopt some algorithm, such as the technique of Fourier transform, to afresh compose signals and analyze them. When the Fourier analysis method is used in the on-line monitoring of dielectric loss,the non-synchronization between the data sampling rate and the frequency of the power network can cause frequency leakage, which leads to inferior precision,even remarkable error.In order to improve the accuracy , this paper offers an algorithm that utilize of window functions and interpolation to track the change of power network frequency and correct the measured signal parameter, which taken by Fourier harmonic analysis. The computing quantity of this algorithm is fewer, and the influence of harmonic on this algorithm is smaller. The result of data signal trial proves that it has higher accuracy , especially the Blackman-Harris window.As the core of dielectric loss detection , data is gathered by Single-Chip, then the computer receives the data via communication with RS485. Virtual instrument software,LabVIEW,deals with these data and calculates important parameter, including frequency, phase and amplitude. The function of system consist of dielectric loss detection, display, query, alarm and remote diagnosis by experts.