Research On On-line Monitoring System Of Interturn Short Circuit Of Reactor Based On Magnetic Field Detection

As one of the important electrical equipment in the power system,the dry-type aircore reactor plays an important role in compensating reactive power,maintaining system voltage,limiting short-circuit current and so on.In actual power grid,dry-type air-core reactor is usually installed on the outdoor,because the factors,such as work environment and system overvoltage reactor often occur because of turn-to-turn insulation defects and partial discharge and cause of interturn short circuit fault,causing fire accident,the serious influence the safe and stable operation of power system.In view of the lack of sensitivity and accuracy of current detection methods for interturn short circuit of reactors,a new detection method for interturn short circuit magnetic field based on wavelet analysis is proposed in this paper according to the variation characteristics of interturn short circuit magnetic field distribution.Firstly,the principle of magnetic field detection method was analyzed and introduced according to the interturn short circuit model of air-core reactor.Combined with the structural characteristics of air-core reactor,a CKDK-10/100-5 dry-type air-core series reactor was taken as a calculation example,and the two-dimensional finite element model of the reactor was established using ANSYS Maxwell software.The magnetic field-circuit coupling analysis method is used to simulate the current change and magnetic field distribution change rule when the reactor is wrapped in different positions and short circuit between different turns,which provides a theoretical basis for the online monitoring method of the reactor short circuit between turns based on magnetic field detection.In view of the strong electromagnetic interference in the reactor working environment and the influence of system sampling error,according to the change characteristics of detecting coil signal during the reactor fault development,the wavelet transform is proposed to analyze and process the detecting coil signal.Based on the principle that the relative ratio between useful signal and noise is the largest in the wavelet decomposition coefficient of signal,the optimal wavelet basis function is selected layer by layer,and the threshold denoising of signal is carried out by wavelet transform,and the fault characteristic quantity is extracted so as to realize timely and accurate judgment of the development period of short circuit between turns of reactor and avoid further development of fault.The wavelet analysis of detection coil signal is carried out by MATLAB.The results show that the wavelet denoising algorithm can suppress the interference noise better than the traditional wavelet algorithm,so as to identify the short-circuit fault signals between turns at different positions of the reactor quickly and reliably.Finally,on the basis of theoretical research and simulation analysis,this paper designs and makes an online monitoring system for the interturn short circuit of the reactor,analyzes the hardware circuit functions of each module in detail,and designs the corresponding circuit schematic diagram.Meanwhile,the software program corresponding to the on-line monitoring system is developed based on modular programming.The experimental platform of online monitoring system for interturn short circuit of reactor was built under laboratory conditions,and experimental research on interturn short circuit of reactor at different positions was carried out.Through comparative analysis of experimental data and simulation results,it can be seen that the magnetic field detection method proposed in this paper based on wavelet analysis has high sensitivity and reliability.It proves the reasonableness of prototype design of online monitoring system for interturn short circuit of reactor,and provides a new idea for the development of online monitoring system for interturn short circuit of reactor.