基于广义解调和广义S变换的时频域去噪方法  

Time-frequency Denoising Method Based on Generalized Demodulation and Generalized S-transform

作　　者：柏晨[1] 肖泽龙[1] 许建中[1] 

机构地区：[1]南京理工大学电子工程与光电技术学院,江苏南京210094

出　　处：《探测与控制学报》2011年第4期28-33,共6页Journal of Detection & Control

摘　　要：毫米波干涉仪回波信号的信噪比对膛内高速运动目标参数的测试精度有重要影响,传统的时域或频域去噪方法经常受到使用条件的限制,为此提出一种基于广义解调和广义S变换的时频域去噪方法。首先估计信号的相位函数对其进行广义解调和广义S变换,在时频域构造有效的滤波器进行降噪处理,随后对去噪信号进行广义S逆变换和逆广义解调并在此基础之上得到目标的运动曲线。基于此方法的二维滤波器参数完全与时间无关,极大地降低了时频滤波器的构造难度。仿真和实测数据分析结果表明:该算法与希尔伯特黄等方法相比能够有效提高信噪比并改善测试精度。The signal to noise ratio of echo of the millimeter-wave interferometer has a significant effect on the testing precision of high speed moving objects parameters,but traditional denoising methods based on time-domain or frequency-domain are often restricted by practical conditions.A denoising algorithm based on generalized demodulation and generalized S-transform was proposed in this paper.The signal was generally demodulated and S-transformed with an estimating phase,and denoised with a effective filter in the time-frequency domain.Finally,with the anti-generalized demodulation and anti-generalized S-transform,the mathematic moving function of the target was calculated based on this denoising signal.The difficulty of creating the 2-D filter was greatly diminished for its parameters had no relation to time factors.Simulation results and testing data analysis showed that,comparing with Hilbert-Huang transform and other algorithms,the signal to noise ratio and testing precision were promoted effectively by this method.

关 键 词：毫米波干涉仪 时频分析 广义解调 广义S变换 

分 类 号：TN911.7[电子电信—通信与信息系统]