Research On Key Technology Of Collective Capacitor Intelligent Components

With the rapid development of smart substation, requirements of intelligent detection for high-voltage electrical equipment gradually increase. Intelligent components can be real-time monitoring the running state of the equipments, when something be in the problem, workers can make equipment maintenance in time, avoiding the losses caused by sudden blackouts. In view of the current hot spots of intelligent components, this thesis focuses on the measurement algorithm of dielectric loss angle of collective capacitor intelligent components on-line monitoring module.Dielectric loss angle tanδ is one of the basic indicators reflecting the capacitor insulation condition. It can sensitively find out many kinds of the insulation defects of high-voltage capacitor. Such as entire insulation humidity, insulation degradation, the overall defects and the local defects of the Small size equipment. So tanδ is of great significance for the insulation diagnosis of power capacitor.After comparing the harmonic analysis method, high-order sine fitting method and related-function method, in numerous traditional algorithms, this thesis chooses the harmonic analysis method. And adopting four third-orders Nuttall window with double spectral lines interpolation method to make it improved, so as to be the test algorithm of on-line monitoring of dielectric loss angle. This thesis builds the simulation model of the circuit in the Matlab, simulates the normal and fault running state of capacitor respectively, calculates the dielectric loss angle of each case and contrasts with standard values when there are frequency fluctuations, harmonic component and white noise interference in the system. And then, use wavelet transform algorithm to eliminate the Gaussian white noise in the current and voltage signal. At last, the author carry out the field test, builds the model of the normal and fault running state of capacitor, calculates the dielectric loss angle when there are frequency fluctuations and harmonic component in the system. So as to verify the feasibility that harmonic analysis method whether can be used for the on-line monitoring of dielectric loss angle.