摘要
对磁法探测水下目标数据进行去噪处理,可以提高地质探测的精度和效率。进行磁法探测水下目标数据去噪时,应获取磁法探测水下目标数据信号和周围介质间信号之间的特征差异,利用上述差异对目标数据信号进行重建,剔除了全部干扰噪声,但是传统方法将最大化RN分量与标准正态分布的对数似然完成目标数据去噪,无法辨别目标数据和周围介质间信号的特征差异,存在目标数据去噪误差大的问题。提出一种采用稀疏分解的磁法探测水下目标数据去噪方法。上述方法先融合稀疏分解的基本原理比较全部数据信号时域中的延迟性,依据频域幅值大于周围介质间信号幅值的特点,得到能体现两者之间显著差异的特征参数,利用反映水下目标数据信号主要结构的模态对信号进行部分重建,对磁法探测水下目标数据中的噪声进行去除。仿真结果表明,所提方法可以有效去除磁法探测水下目标中干扰噪声,为在水域进行磁法探测提供了科学依据。
A data de-noising method of underwater target detection via magnetic method is proposed by using sparse decomposition. Firstly, the delays in all time domains of data signals are compared through sparse decomposi- tion. Then the feature parameter reflecting the both significant difference is obtained according to the characteristic that frequency domain amplitude is greater than signal amplitude among surrounding media. Finally, the noise in un- derwater target data is eliminated via magnetic method by using the modality reflecting signal primary structure of un- derwater target data to rebuild signal partly. The simulation results show that the proposed method can eliminate the noise in underwater target data.
出处
《计算机仿真》
北大核心
2017年第5期197-200,共4页
Computer Simulation
关键词
磁法探测
水下目标数据
去噪
Magnetic detection
Under-water target data
De-noising
