Research On The Intelligent Active Noise Control System Of Large Power Transformers

With the progress of technology and the development of the society, noise pollution have attracted great attention of people. In recent years, many high voltage transformers are gradually close to the residential area due to the expansion of urban areas and the power grids upgrading, which does harm to people’s work and life.This paper selects a 110kV transformer in the suburbs of a city which operates normally, and collects its noise signals. Through analyzing the noise signals, it is found that transformer noise energy is concentrated in the low frequency region and the noise is higher than other frequencies at 100Hz and frequency multiplication. The traditional measures of transformer noise control are mainly based on sound insulation, sound absorption or sound elimination. But their results are not satisfactory, and cost too much with installing inconveniently. With the development of power electronic technology, the active noise control based on Huyghens principle has obtained considerable development. It is characterized as small equipment and low cost, while effect on low-frequency noise control. So the technology is appropriate to the noise control of high voltage transformers.The core control part of active noise control system is adaptive filter, which can make the system respond to the changes of noise adaptively. The common control algorithm of adaptive filter is LMS algorithm. This paper analyzes algorithm performance in theory. Based on LabVIEW, this paper also designs an adaptive filter to study the factors affecting its convergence rate and steady state error. It is found that the effects of convergence coefficient on the rate of convergence and steady state error are contradictory. To solve this problem, this paper presents a variable-step LMS algorithm. Through the simulation, the algorithm can improve the steady state error with high convergence rate.Based on simulation result, this paper designs an adaptive active noise control of large power transformer with adaptive filter at the core. Its hardware part mainly includes gathering module, control module and output module. The gathering module collects primary signals and error signals, and completes the function of A/D conversion. The control module takes computer as its carrier, adjusts output signals according to the changes of input signals. The output module changes the digital signals into analog signals, and produces secondary noise. The major part of software is adaptive filter and its control algorithm, with data acquisition module and signal output module are responsibility for the communication of computer and instruments. Otherwise this paper adds data analysis module, data memory module, historical data retrieval module and signal processing module, which enrich the functions of software greatly.In the lab environment, this paper performs an experiment with the recording of power transformer as the primary noise. The noise was decreased by 5～10 decibels generally in position of error sensor, and the sound energy declined by 68～90 percent, which proves the effect of active noise control system.