Design And Application Of Noise Measurement And Analysis System For Power Transformer In Smart Grid

In recent years, with the increasing demand for power quality and security of power network, the development speed of smart grid is more and more rapidly. Intelligent substation as a transit hub in the process of power transmission, plays an important role, and whether the power transformer, which is the main equipment in intelligent substation, can reliably run decides safety of the entire power grid.The running state of transformer is closely related to the transformer’s noise. This paper mainly designs a set of noise measurement and analysis system, including collecting the noise of transformer、calculating the sound pressure level and observating and analyzing the noise characteristics of the transformer to achieve the purpose of real-time monitoring the running state of the transformer. In this paper, the design of the system is mainly divided into three parts: the design of the noise measurement and calculation platform、the calculation of the noise signal and the analysis of the noise signal.Firstly, the overall design of the noise measurement and analysis platform is carried out,including the design of hardware selection and software task flow. Hardware design including noise data acquisition module, processing and control module, storage module, and communication module. Software architecture mainly includes controlling the data acquisition process by FPGA and designing CPU task operation flow.Secondly, designed the A、C、Z frequency weighted network based on the IIR digital filter and through the MATLAB simulation to obtain its error in line with the 1 level tolerance standard. It is fully applicable to sound pressure level’s high precision calculation of the power transformer’s noise. The real-time fast fourier transform algorithm is applied to the calculation of the spectrum of the noise. Its computation is less than the common fast fourier transform. By observing the sound pressure level and the spectrum value of the transformer noise signal, it can achieve the purpose of real-time monitoring of the running state of the transformer.Thirdly, carried out the noise signal’s further spectral analysis, and used improvement of 1 / 3 octave method to express FFT, that is using the maximum spectrum value of a frequency band showed spectrum value of the band and its feasibility was verified by the simulation of MATLAB. Aiming at the phenomenon of spectral leakage of FFT, reduce the error by adding window function. Then by compensating the attenuation value caused by adding the window function, the accuracy of FFT analysis is improved. Aiming at the defects of FFT in spectrum analysis, wavelet transform is used to analyze the frequency spectrum of the noise signal, and analysed the detection of noise signal’s frequency component and detection of the noise signal singularity by wavelet analysis. Wavelet analysis can not only analyze the frequency domain’s component, but also analyze the time domain’s characteristics of the signal, which provides an effective analysis method for the fault analysis of power transformer noise signal.Finally, the field noise signal is collected by the noise measurement and analysis platform designed in this paper. The field signal is calculated and analyzed according to the algorithm introduced in this paper, which verified the practicability of the algorithm. The system is fully applicable to the calculation and analysis of transformer noise’s characteristics.