【作者】 闫士杰； 王旭； 黄延龙； 荆艳强；

【Author】 YanShijie WangXu Huang Yanlong JingYanqiang (College Of Information Science And Engineering Norththeast University ShenYang 110004 China)

【机构】 东北大学信息学院；

【摘要】 本文提出了一种基于正交小波变换和自适应学习算法相结合的噪声抑制方法。该方法通过正交小波变换将原始信号分解到若干不同的频带上,利用基于梯度的自适应学习算法求取最佳阈值,这样在消噪过程中可以得到最小均方误差,达到或接近最优的消噪效果,克服了硬阈值和软阈值方法的缺点。该方法应用在微型燃机系统中的排气温度检测,通过仿真,验证了小波阈值消噪以及自适应阈值消噪算法的优越性能。更多还原

【Abstract】 A novel noise suppression method was proposed,which is based on orthogonal wavelet transform and adaptive learning algorithm.In the algorithm original signal was changed into a number of small bands by mean of orthogonal wavelet transform and optimal threshold was gotten utilizing adaptive learning algorithm based on gradient,so the minimum mean square error can be gotten and optimal effect of noise elimination can be achieved or closed to.The shortcoming of threshold of hard and soft threshold method was overcome.This algorithm was already applied to detect exhaust temperature signal in Microturbine system.Through simulation,excellent performance of the algorithm was tested.更多还原