-
本文提出一种基于变分模式分解(VMD)与独立元分析(ICA)相结合的水下信号处理方法。该方法运用变分模式分解(VMD)将一组回波信号分解为多组按照其频率高低顺序排列的本征模态信号。然后,将这些模态信号作为独立元分析(ICA)的观测矩阵,以确保分离所得目标回波信号的完整性。该方法提出将分解所得的各层模态信号与原信号进行相关性和信杂比比较,以确定其分解层数。应用独立元分析方法(ICA)对散射与目标回波进行分离,从而恢复强散射水体中的微弱目标回波,大大提高其测距精度。进行不同衰减长度水体的532nm调频连续光水下测距实验。经实验验证,该信号处理方法在激光输出功率2.3 W时,成功实现对9个衰减长度内目标的测量,使用算法将测距精度由16 cm提升至5cm以内。This paper proposes an underwater signal processing method based on the combination of Variational Mode Decomposition (VMD) and Independent Component Analysis (ICA). In this method, VMD is used to decompose a set of echo signals into groups of eigenmodal signals arranged in order of their frequency. These modal signals are then used as an observation matrix for ICA to ensure the integrity of the separated target echo signals. In this method, the correlation between the decomposed modal signals and the original signal is used to select the signals which are used as input matrix rows for ICA. The signal-to-clutter ratio is applied to determine the number of decomposed layers. The ICA is used to separate the scattering and target echoes, so as to recover the weak target echoes in the strongly scattered water and greatly improve the ranging accuracy. 532 nm intensity-modulated continuous wave (CW) laser was used in underwater ranging experiments. The attenuation coefficient of the water was changed by adding Mg(OH)2, ranging experiments were carried out at different attenuation lengths. The experimental results show that the signal processing method can successfully measure targets within 9 AL when the laser output power is 2.3 W, and the ranging accuracy is improved from 16 cm to less than 5 cm by using the algorithm. The proposed method can be applied to underwater lidar application in turbid water and long distance ranging underwater where the scattering dominates the echoes.
-
[1] Liang J, Peng S P, Zheng J 2014 Computer Engineering and Applications. 504 7. [梁喆, 彭苏萍, 郑晶.2014计算机工程与应用504 7]
[2] Battista B M, Knapp C C, McGee T, Goebel V 2007 Geophysics 722 29.
[3] Jia R S, Zhao T P, Sun H H 2015 Chinese J Geophys. 583 1013. [贾瑞生, 赵同彬,孙红梅2015地球物理学报583 1013]
[4] Li Y, Peng J L, Ma H T 2013 Chinese J Geophys. 562 626. [李月,彭蛟龙,马海涛 2013地球物理学报562 626]
[5] Huang N E, Shen Z, Long S R, Wu M L, Shi H H, Zheng Q N, Yen N C, Tung C C, and Liu H H 1998 Roy. Soc. A-Math. Phy. 4541971903.
[6] Zheng Z G 2010 Beijing: Meteorological Publishing House 83. [郑祖光2010北 京:气象出版社83]
[7] Hu A J, Sun J J, Xiang L 2011 Vibration, Testing and Diagnosis 314 429. [胡爱军, 孙敬敬,向玲2011振动、测试与诊断314 429]
[8] Konstantin D, Dominique Z 2014 IEEE Transactions on Signal Processing 623 531.
[9] Dey P, Satija U, Ramkumar B 2015 Annual IEEE India Conference (INDICON) New Delhi.
[10] Gao Y F, Zhu Y L, Yan H 2016 Transaction of China Electrotechnical Society 311 24.[高艳丰,朱永利,闫红艳, 2016电工技术学报311 24]
[11] Ma Z Q, Li Y C, Liu Z 2016 vibration and shock. 3513134. [马增强,李亚超, 刘政2016振动与冲击3513134]
[12] Xie P, Yang F M, Li X X 2016 Acta Phys Sinica. 6511118701. [谢平,杨芳梅, 李欣欣2016物理学报6511118701]
[13] Tang G J, Wang X L, 2015 Journal of Xi'an Jiao tong University. 495 73. [唐贵基, 王晓龙2015西安交通大学学报495 73]
[14] P Comon 1994 SignalProcess. 363 287.
[15] Laubach M, Shuler M, and M A L 1999 Nicolelis Neurosci.Meth. 941 141.
[16] Bell A J and Sejnowski T J 1995 Neurosci.Methods. 76 1129.
[17] Hyvarnen A and Oja E 2000 NeuralNetworks. 134-5411.
[18] Hyvarnen A and Oja E 1997 NeuralComput. 97 1483.
[19] Vrabie V D, Mars I J, and Lacoume J L 2004 SignalProcess. 843 645.
[20] Nian R, Liu F, and Bo B 2013 Sensors 137 9104.
[21] Illig D W, Jemison W D, and Mullen L J 2016 Appl.Opt. 5531 C25.
[22] Prasad R, Deo C R, and Li Y 2018 Geoderma330136.
[23] Colominas M A, Schlotthauer G, and Torres M E 2014 Biomed. SignalProces. 14, 19.
[24] Yu Y, D J Yu, and Cheng J S 2006 SoundVib. 2941-2269.
[25] Kim D and Kim S K 2012 Behev.Res.Methods 444 1239–1243.
[26] Li K, Yang S H, Liao Y Q, Lin X T, Wang X, Zhang J Y, and Li Z 2020 IEEE Photonics J. 125 1503811.
计量
- 文章访问数: 203
- PDF下载量: 0
- 被引次数: 0