基于EMD与SVM的仿生机器鱼人工侧线智能探测方法

刘 钰; 胡 桥; 赵振轶; 魏 昶

doi:10.11993/j.issn.1673-1948.2019.02.008

基于EMD与SVM的仿生机器鱼人工侧线智能探测方法

doi: 10.11993/j.issn.1673-1948.2019.02.008

刘钰^1,3,
胡桥^1,2,3,
赵振轶^1,3,
魏昶^1,3

1.
西安交通大学机械工程学院, 陕西西安, 710049
2.
西安交通大学机械制造系统工程国家重点实验室,陕西西安, 710049
3.
西安交通大学陕西省智能机器人重点实验室, 陕西西安, 710049

基金项目: 国家自然科学基金重大项目(61890961); 装备预研领域基金项目(61404160503); 中央高校基本科研业务费(国防重大项目培育xjjgf2018005); 陕西省重点研发计划重点项目资助(2018ZDXM-GY-111)

详细信息

作者简介:
刘钰(1994-), 男, 在读博士, 主要研究方向为水下仿生智能感知.

中图分类号: TP242; TB566
计量
- 文章访问数: 578
- HTML全文浏览量: 21
- PDF下载量: 339
- 被引次数: 0
出版历程
- 收稿日期: 2016-11-19
- 修回日期: 2016-12-18
- 刊出日期: 2019-04-30

Intelligent Detection of Artificial Lateral Line for Biomimetic Robotic Fish Based on EMD and SVM

LIU Yu^1,3,
HU Qiao^1,2,3,
ZHAO Zhen-yi^1,3,
WEI Chang^1,3

1.
School of Mechanical Engineering, Xi’an Jiao Tong University, Xi’an 710049, China
2.
State Key Laboratory of Manufacturing Systems Engineering, Xi’an Jiao Tong University, Xi’an 710049, China
3.
Shaanxi Key Laboratory of Intelligent Robots, Xi’an Jiao Tong University, Xi’an 710049, China

摘要

摘要: 针对水下声学感知与信息交互系统常常受到混响或多途效应干扰, 而光学传感也容易受到水质浑浊等环境制约的现状, 为了解决仿生机器鱼在复杂水下环境干扰下的目标精确探测问题, 文中提出一种基于仿生人工侧线(ALL)的水下运动目标智能探测方法。首先, 通过经验模式分解(EMD)将ALL系统接收到的原始信号分解为不同的本征模态函数(IMFs), 从而将有用信号与干扰进行分离; 然后, 在水中放置不同频率的振动物体模拟不同鱼类, 通过IMFs中的特征频率进行水下运动目标的识别; 最后, 将IMFs的能量值频谱信息作为支持向量机(SVMs)的输入, 从而智能识别水下运动物体的方位。将这一探测方法与ALL结合并进行水池实验验证, 结果表明, 文中所提出的方法相比于传统的如快速傅里叶变换(FFT)和神经网络等方法具有更好的探测性能。
- 仿生机器鱼 /
- 人工侧线 /
- 经验模式分解 /
- 支持向量机 /
- 智能探测
Abstract: In view of the problems that underwater acoustic perception and information interaction system are often disturbed by reverberation or multipath effects, and the optical sensing is easily affected by water turbidity, a novel intelligent detection method of underwater moving target based on artificial lateral line(ALL) is proposed to accurately detect target in complicated underwater interferences for a biomimetic robotic fish. Firstly, the original signals received by the ALL system are decomposed into different intrinsic mode functions(IMFs) via empirical mode decomposition(EMD) to separate target signal from interferences. Secondly, the characteristic frequency of the vibrating target representing different fish is obtained from the IMFs to detect the underwater moving target. Finally, the power spectrums of the IMFs are input into the support vector machines(SVMs) to recognize the azimuth of underwater moving target intelligently. Water tank tests with combination of the proposed method with ALL detection are conducted, and the results show that the proposed method has better detection performance than the traditional methods, such as fast Fourier transform(FFT) and neural network.
- biomimetic robotic fish /
- artificial lateral line(ALL) /
- empirical mode decomposition(EMD) /
- support vector machine(SVM) /
- intelligent detection

HTML全文

参考文献(12)

[1]	[1] Du R, Li Z, Youcef-Toumi K, et al. Robot Fish: Bioin-spired Fishlike Underwater Robots[J]. ?Underwater Technology, 2017, 34(3): 143-145.
[2]	Bleckmann H, Zelick R. Lateral Line System of Fish[J]. Integrative Zoology, 2006, 25(1): 411-453.
[3]	Olt J, Stuart L J, Marcotti W. In Vivo and in Vitro Bio-physical Properties of Hair Cells From the Lateral Line and Inner Ear of Developing and Adult Zebrafish[J]. The Journal of Physiology, 2014, 592(10): 2041-2058.
[4]	Dabiri J O. Biomechanics: How Fish Feel the Flow[J]. Nature, 2017, 547(7664): 406-407.
[5]	Oteiza P, Odstrcil I, Lauder G, et al. A Novel Mechanism for Mechanosensory-based Rheotaxis in Larval Zebrafish[J]. Nature, 2017, 547(7664): 445.
[6]	Wang W, Li Y, Zhang X, et al. Speed Evaluation of a Freely Swimming Robotic Fish with an Artificial Lateral Line[C]//IEEE International Conference on Robotics and Automation. Stockholm, Sweden: IEEE, 2016.
[7]	Ahrari A, Lei H, Sharif M, et al. Reliable Underwater Dipole Source Characterization in Three-dimensional Space by an Optimally Designed Artificial Lateral Line System[J]. Bioinspiration & Biomimetics, 2017, 12(3): 036010.
[8]	Dagamseh A M K, Lammerink T S J, Kolster M L, et al. Dipole-source Localization Using Biomimetic Flow-sensor Arrays Positioned as Lateral-line System[J]. Sensors & Actuators A Physical, 2010, 162(2): 355-360.
[9]	Abdulsadda A T, Tan X. Underwater Source Localization Using an IPMC-based Artificial Lateral Line[C]//IEEE International Conference on Robotics and Automation. Shanghai, China: IEEE, 2011: 2719-2724.
[10]	Ahrari A, Lei H, Sharif M A, et al. Design Optimization of an Artificial Lateral Line System Incorporating Flow and Sensor Uncertainties[J]. Engineering Optimization, 2016, 49(2): 328-344.
[11]	Kottapalli A G P, Asadnia M, Triantafyllou M S, et al. Biomimetic Microsensors Inspired by Marine Life[M]. Switzerland: Springer, 2017.
[12]	Huang N E , Shen Z , Long S R , et al. The Empirical Mode Decomposition and the Hilbert Spectrum for Nonlinear and Non-stationary Time Series Analysis[J]. Proceedings A, 1998, 454(1971): 903-995.