基于多频陷波滤波和信号功率变参数的变结构FXLMS自适应变压器有源降噪算法

发布时间：2018-03-15 07:19

本文选题：电力变压器噪声　切入点：有源降噪系统　出处：《中国电机工程学报》2017年22期 　论文类型：期刊论文

【摘要】：为了改善变压器有源降噪系统的性能,提出了基于多频陷波滤波和信号功率变参数的变结构滤波-X最小均方(filter-X least mean square,FXLMS)自适应有源降噪算法。该算法以变结构FXLMS算法为核心,针对变压器噪声的特点,采用多频陷波滤波技术实现了系统无参考传声器的设计,降低硬件成本并节约占地面积。采用随机噪声法对系统次级通道进行在线建模,并提出了"实时建模,延时更新"的策略,以及基于信号功率的变建模收敛系数和变随机噪声幅值相结合的方法,以降低引入的随机噪声幅值,并保证建模的准确性。同时,基于信号功率实现了对降噪收敛系数的自适应调节。仿真和理论分析表明,该算法可加快系统的收敛速度,提高系统的稳定性和降噪量。设计的软件平台与硬件设备相结合的变压器有源降噪系统,在额定运行的50MVA变压器降噪实验中,实现了误差传声器处10~18d B的降噪效果,平均声能量密度降低90.01%~98.41%,验证了算法对实际电力变压器噪声的降噪有效性。
[Abstract]:In order to improve the performance of transformer active noise reduction system, an adaptive active noise reduction algorithm based on multi-frequency notch filter and variable signal power variable parameter is proposed, which is based on variable structure FXLMS algorithm. According to the characteristics of transformer noise, the multi-frequency notch filter technique is used to realize the design of the system without reference microphone, which reduces the hardware cost and saves the area. The stochastic noise method is used to model the secondary channel of the system online. The strategy of "real-time modeling, delay updating" and the combination of variable modeling convergence coefficient and variable random noise amplitude based on signal power are put forward to reduce the amplitude of the introduced random noise and ensure the accuracy of modeling. Based on the signal power, the adaptive adjustment of the noise reduction convergence coefficient is realized. The simulation and theoretical analysis show that the algorithm can accelerate the convergence speed of the system. To improve the stability and noise reduction of the system. The designed active noise reduction system of transformer based on software platform and hardware equipment is used in the noise reduction experiment of 50MVA transformer with rated operation, and the noise reduction effect of 10 ~ 18dB at the error microphone is realized. The average acoustic energy density is reduced by 90.01 and 98.41, which verifies the effectiveness of the algorithm in noise reduction of practical power transformers.
【作者单位】：山东大学电气工程学院;
【基金】：国家自然科学基金项目(51207084) 山东省优秀中青年科学家科研奖励基金计划(BS2013NJ024)~~
【分类号】：TM41

【相似文献】