Research On Anti-interference Technology Of Active Noise Control For Power Transformer

With the rapid development of China's economy and urbanization,the high voltage and large capacity substations which were originally located in the suburbs gradually go into the residential areas and urban centers.With the public's improvement of the acoustic environmental quality requirement,the noise pollution of transformer is widely concerned by the society.Therefore,It is very meaningful to effectively control the noise pollution of the transformer.By optimizing the manufacturing process to reduce the transformer noise is a systematic project,which not only needs to increase the huge investment,but also has many technical problems that have not yet been resolved fundamentally up.The traditional passive noise control technology,such as enclosures,barriers,silencers and so on,are relatively large,costly,and ineffective at the low frequency noise of the transformer.Therefore,the research of active noise control technology of transformer has the practically applied value.In this paper,the influence of impulsive interference and secondary acoustic feedback on the stability of active noise control system for the transformer are studied,and the anti-interference ability of this system can be enhanced by improving the algorithm and system structureThe impulsive interference signals acquired from acoustic sensors has a great effect on the stability of ANC system.The stability of controller under the impulsive disturbance in reference signal and error signal are analyzed in detail,ie,impulsive interference deteriorate ANC system by deviating weights of ANC controller from its converged values.In the worst case,it may cause adaptation divergence.Based on the analysis and summary of research at home and abroad,normalized filtered-x p.law LMS algorithm is proposed with minimized squared ?-law of speech compression transformation of error signal as its cost function.And reference signal is normalized to improve robustness of the algorithm.Filter coefficients are updated with a gradient weighting function,which can be controlled with compression parameter to adapt characteristics of sound field of substations.Contrast simulation results demonstrate the effectiveness of the proposed algorithm.In order to avoid the problem of secondary acoustic feedback,this paper studies the feasibility of the application of the internal reference signal to the active noise control of the transformer.Firstly,the characteristics of the parallel and direct multi frequency ANC system based on the internal reference signal are compared and studied,and the active noise control system for transformer is constructed by using the parallel structure.Then,the noise reduction formula is derived which take into account the effect of frequency mismatch between the transformer noise and the internal reference signal.And it can be concluded that when there is a small range of frequency mismatch,the noise reduction can be maintained by choosing a larger iteration step.Finally,the simulation results show the effectiveness of the internal reference signal method applied to the active noise control system of transformer.The convergence speed and thus the noise reduction of every subsystem of the parallel multi frequency ANC system will be degraded because of using the same error signal to update the weight coefficient.This paper develop a complete parallel ANC system that use the bandpass filterbank to split the error signal,and uses the individual error signal to update the corresponding subsystem.In the new system,the adaptive updating process of every subsystem controller does not affect each other.The convergence speed and convergence precision of the system before and after the modification is compared,and the practicability of the new system is verified.