Research On Noise Characteristics And Active Noise Reduction System Of Uhv Transformer

With the increasing demand for social electricity,the number of UHV substations increases year by year,and the noise generated by its viberation has an increasingly serious influence on the work and daily life of the surrounding residents.In this paper,the 1000 kV UHV Transformer in Changzhi substation is selected for the study.The methods of sound pressure,sound intensity and vibration are used to measure and analyze the body noise.The analysis results show that the frequency of the body noise of UHV Transformer is below 600 Hz,and is mainly distributed in 100 Hz and its integer times.The noise reduction measures such as sound insulation and sound absorption are not obvious for low frequency noise,and changing the material and structure of the core of the transformer will also increase manufacturing costs.With the development of digital signal processing technology,active noise reduction technique has attracted increasing attention,it has advantages of small system,low cost,low frequency effect,etc,and apply to noise control of UHV Transformer.Active noise reduction technology is mainly based on acoustic interference principle,by adjusting the amplitude and phase of the secondary sound source,then interference with primary sound source to achieve noise reduction in the target area.Therefore,it is very important for the system design of active noise reduction to study the sound propagation characteristics of the primary and secondary sources and the distribution of the sound field caused by the coherent effects of the acoustic wave.In the laboratory,the loudspeaker is used to simulate the transformer noise in the experiment of active noise reduction,and the simple sound source is chosen as the equivalent model of the loudspeaker,which results in inaccurate results due to the large model error.Since the speaker of a paper basin can be equivalent to a dipole source at low frequencies,dipole acoustic source is chosen as the equivalent model of the loudspeaker in this paper.In order to compare and analyze,the sound field interference characteristics of a loudspeaker equivalent to a point source and a dipole source model are simulated and experimentally verified in this paper.The results show that the dipole acoustic source model is more in line with the actual radiation characteristics of the speaker than the point source model.In order to achieve the best effect of noise reduction,active noise reduction system must have adaptive function.Therefore,this paper selects the least mean square error(LMS)adaptive filtering algorithm and build a simulation system.Simulation results show that the algorithm is effective in noise reduction.The effect of convergence coefficient and filter length on the performance of LMS adaptive filtering algorithm is also studied.The simulation results show that increasing the convergence speed or increasing the convergence coefficient of filter coefficient will both accelerate the adaptive algorithm,and increase the amount of noise reduction.but when the value of convergence coefficient and the filter coefficient exceeds a certain threshold,the steady-state error will increase,and the system will be turbulent until the final divergence.Therefore,in order to obtain better noise reduction effect,the numerical values of each influence factor should be set reasonably,and the simulation results will provide data support for the selection of the experimental parameters of the active noise reduction.Finally,the hardware and software of active noise control system is designed,and the noise reduction experiment is carried out,achieved about 7dB noise reduction.verified the effective noise reduction of the adaptive filtering algorithm and active noise reduction system in this paper.