最小二乘权值平滑滤波技术在核磁共振信号处理中的应用  

Least-square weighted smoothing filter technology applied in magnetic resonance sounding signal processing

作　　者：孙淑琴[1,2] 刘骏妍 蒋川东[1,2] 孟庆云[1,2] 林君[1,2] 彭良玉[1,2] 

机构地区：[1]吉林大学仪器科学与电气工程学院,长春130026 [2]吉林大学地球信息探测仪器教育部重点实验室,长春130026

出　　处：《吉林大学学报（工学版）》2016年第3期985-995,共11页Journal of Jilin University:Engineering and Technology Edition

基　　金：国家重大科学仪器设备开发专项项目(2011YQ030133)

摘　　要：采用最小二乘权值平滑滤波法对核磁共振地下水探测仪接收到的信号进行滤波处理,并与滑动平均权值滤波方法进行比较。仿真试验结果表明:利用最小二乘权值平滑滤波法能够取得更准确的核磁共振(Magnetic resonance sounding,MRS)信号特征参数,信噪比也得到了提高;最小二乘权值平滑滤波法更适合于随机噪声和尖峰噪声的去除;利用最小二乘权值平滑滤波法进行去噪时要选择合适的权值。其中MRS信号特征参数的拟合误差能够控制在1%以内,且信号曲线的平滑度得到了很大的改善。同时实测数据的滤波处理进一步证明了该方法的可行性。Least Square Weighted Smoothing Filter（LSW-SF）method is used to filter Magnetic Resonance Sound（MRS）signals,which is compared with the moving average filter method.Results of a large number of simulation experiments show the following characteristics.First,the more accurate characteristic parameters of MRS can be obtained using the LSW-SF method,and the Signal to Noise Ratio（SNR）is also improved.Second,the LSW-SF method is more suitable for random noise and rush noise removal.Third,appropriate weights should be chosen when the LSW-SF method is used to wipe out noise.Among them fitting error of MRS signal characteristic parameters can be controlled within 1%,and the smoothness of the signal curve is greatly improved.Meanwhile,thefilter processing of the measured data indicates this method is effective.

关 键 词：电气工程 核磁共振 数据滤波 滑动平均 最小二乘 

分 类 号：TN911[电子电信—通信与信息系统]