A GIM-Based Biomimetic Learning Approach for Motion Generation of a Multi-Joint Robotic Fish

doi:10.1016/S1672-6529(13)60237-1

摘要/Abstract

关键词: robotic fish, locomotion, biomimetic learning, GIM

Abstract:

In this paper, we propose a biomimetic learning approach for motion generation of a multi-joint robotic fish. Based on a multi-joint robotic fish model, two basic Carangiform swimming patterns, namely “cruise” and “C sharp turning”, are extracted as training samples from the observations of real fish swimming. A General Internal Model (GIM), which is an imitation of Central Pattern Generator (CPG) in nerve systems, is adopted to learn and to regenerate coordinated fish behaviors. By virtue of the universal function approximation ability and the temporal/spatial scalabilities of GIM, the proposed learning approach is able to generate the same or similar fish swimming patterns by tuning two parameters. The learned swimming patterns are implemented on a multi-joint robotic fish in experiments. The experiment results verify the effectiveness of the biomimetic learning approach in generating and modifying locomotion patterns for the robotic fish.

Key words: robotic fish, locomotion, biomimetic learning, GIM

Qinyuan Ren, Jianxin Xu, Lupeng Fan, Xuelei Niu. [J]. J4, 2013, 10(4): 423-433.

Qinyuan Ren, Jianxin Xu, Lupeng Fan, Xuelei Niu. A GIM-Based Biomimetic Learning Approach for Motion Generation of a Multi-Joint Robotic Fish[J]. J4, 2013, 10(4): 423-433.

参考文献

[1] Colgate J, Lynch K M. Mechanics and control of swimming: A review. IEEE Journal of Oceanic Engineering, 2004, 29, 660–673.

[2] Sfakiotakis M, Lane D M, Davies J. Review of fish swim-ming modes for aquatic locomotion. IEEE Journal of Oce-anic Engineering, 1999, 24, 237–252.

[3] Liu J, Hu H. Biological inspiration: from carangiform fish to multi-joint robotic fish. Journal of Bionic Engineering, 2010, 7, 35–48.

[4] Lighthill J L. Mathematical Biofluiddynamics, SIAM, Philadelphia, PA, USA, 1975.

[5] Newman J N, Wu T Y. Hydromechanical aspects of fish swimming. In: Wu T, Brokaw C J, Brennan C (eds). Swim-ming and Flying in Nature, Vol. 2, Plenum Press, New York, USA, 1975, 615–634.

[6] Churchill S W. Viscous Flows: The Practical Use of Theory, Butterworth, London, UK, 1988.

[7] Wolfgang M J, Anderson J M, Grosenbaugh M A, Yue D K, Triantafyllou M S. Near-body flow dynamics in swimming fish. The Journal of Experimental Biology, 1999, 202, 2303–2307.

[8] Weihs D. A hydrodynamical analysis of fish turning ma-noeuvres. Proceedings of the Royal Society B, 1972, 182, 59–72.

[9] Weihs D. The mechanism of rapid starting of slender fish. Biorheology, 1973, 10, 343–350.

[10] Ahlborn B, Harper D, Blake R, Ahlborn D, Cam M. Fish without foot prints. The Journal of Theory of Biology, 1991, 148, 521–533.

[11] Ahlborn B, Chapman S, Stafford R, Blake R W, Harper D. Experimental simulation of the thrust phases of fast-start swimming of fish. The Journal of Theory of Biology, 1997, 200, 2301–2312.

[12] Kelly S D, Murray R M. Modelling efficient pisciform swimming for control. International Journal of Robust and Nonlinear Control, 2000, 10, 217–241.

[13] Kelly S D, Mason R J, Anhalt C T, Murray R M, Burdick J W. Modeling and experimental investigation of carangiform locomotion for control. Proceedings of the American Con-trol Conference, 1998, 1, 1271–1276.

[14] Yu J, Chen E, Wang S, Tan M. Research evolution and analysis of biomimetic robot fish. Journal of Control Theory and Application, 2002, 24, 169–174.

[15] Liu J, Dukes L, Knight R, Hu H. Development of fish-like swimming behavious for an autonomous robotic fish. Pro-ceedings of the Control, University of Bath, UK, 2004, 6–9.

[16] Marder E, Bucher D. Central pattern generators and the control of rhythmic movements. Current Biology, 2001, 11, 986–996.

[17] Hu Y, Tian W, Liang J, Wang T. Learning fish-like swim-ming with a CPG-based locomotion controller. IEEE In-ternational Conference on Intelligent Robots and Systems, San Francisco, CA, USA, 2011, 1863–1868.

[18] Ijspeert A J. A connectionist central pattern generator for the aquatic and terrestrial gaits of a simulated salamander. Bio-logical Cybernetics, 2001, 84, 331–348.

[19] Ijspeert A J. Central pattern generators for locomotion con-trol in animals and robots: A review. Neural Networks, 2008, 21, 642–653.

[20] Xu J X, Ren Q, Niu X. Mimicry of fish swimming patterns in a robotic fish. IEEE International Symposium on Industrial Electronics, Hangzhou, China, 2012, 1274–1279.

[21] Ren Q, Xu J X, Gao W, Niu X. Generation of robotic fish locomotion through biomimetic learning. IEEE/RSJ Inter-national Conference on Intelligent Robotics and System, Vilamoura, Portugal, 2012, 815–821.

[22] Liu J, Hu H. Mimicry of sharp turning behaviours in a ro-botic fish. IEEE International Conference on Robotic and Automation, Barcelona, Spain, 2005, 3318–3323.

[23] Yu J, Wang M, Tan M, Li Y F. Step function based turning maneuvers in biomimetic robotic fish. International Con-ference on Intelligent Robotics and System, Kobe, Japan, 2009, 3431–3436.

[24] Spierts I L, Leeuwen J L. Kinematics and muscle dynamics of C- and S-starts of carp. Journal of Experimental Biology, 1999, 202, 393–406.

[25] Nishikawa I, Hayashi K, Sakakibara K. Complex-valued neuron to describe the dynamics after Hopf bifurcation: An example of CPG model for a biped locomotion. Interna-tional Joint Conference on Neural Networks, Orlando, FL, USA, 2007, 327–332.

[26] Seo K, Chung S J, Slotine E. CPG-based control of a tur-tle-like underwater vehicle. Autonomous Robots, 2010, 28, 247–269.

[27] Xu J X, Wang W. A general internal model approach for motion learning. IEEE Transaction on system, Man, and Cybernetics, 2008, 38, 477–487.

[28] Xu J X, Ren Q, Zhao Q, Niu X. Motion control for a multi-joint robotic fish with the pectoral fins assistance. The 2013 IEEE International Symposium on Industrial Elec-tronics (ISIE), Taipei, 2013, 1–6.