Optimization and control of gait rehabilitation robot

Document Type : Original article

Authors

1 1. Master student in Mechanical Department., Faculty of Engineering, University of Isfahan. Isfahan. Iran

2 2. Biomedical engineering PhD, Assistant Professor of Biomedical Engineering, Faculty of Engineering, University of Isfahan. Isfahan. Iran

3 3. Electrical PhD, Associate Professor of Electrical Engineering, Electrical Engineering Department, University of Isfahan. Isfahan. Iran

4 4. Master student in Mechanical Department, Faculty of Mechanical Engineering, Isfahan University of Technology.

5 5. MD, Cardiolosist, Aadvanced (3D) Echocardiologist, Associate professor Of Isfahan University Of Medical Sciences of occupational therapy, Tehran University of Medical Sciences, Tehran, Iran

Abstract

Background and Aim: In many fields of rehabilitation robotics, desired torque would be obtained by considering dynamics of the actuators. Also, dynamic model of many actuators is hard to approach and it has been neglected in many researches that cause the considerable disadvantages in general results. . The aim of this study is to optimize the performance of gait rehabilitation robot.
Materials and Methods: Kinematic data obtained from 10 patients, including joint angles, angular velocity and angular acceleration that eighty data recorded in a walk cycle and applied as inputs to the neural network. Then, performance of a NARMA-L2 controller for actuators of the biped walking robot is shown. It is noticeable that the controller is learned by LM algorithm and three evolutionary algorithms; PSO, GA and ICA. For controlling the robot walking, two kinds of dc motors are used. These actuators improve control of the biped robot by tracking the required torques for a human walk cycle as a reference model by neural network in two ways; offline and online.
Results: In this research, gait recovery system is introduced that the dynamics of the actuators are considered. Finally, reducing dimensions of the neural network hardware and using the evolutionary algorithms, respectively, reduces costs and increases accuracy of the rehabilitation biped robot based on reducing the minimum mean square error of 20 times after learning of neural network.
Conclusion: This system would be effective in promoting gait recovery. Helping the recovery of patient using robotic system and muscle compensation was the predictive goals of manufacturing of the system. The simple structure of neural network, negligible tracking error and high speed system recognition are the factors that make it the best system recognition in this field.

Keywords

References

  1. Lo H.S, Xie SQ. Exoskeleton robots for upper-limb rehabilitation: State of the art and future prospects. Medical Engineering & Physics 2012; 34(2): 261-268##
  2. Zoss A.B, Kazerooni H, Chu A. Biomechanical design of the Berkeley lower extremity exoskeleton (BLEEX), IEEE/ASME Transactions on Mechatronics 2006; 11(2): 128-13##
  3. Banala S.K, Kim SH, Agrawal SK. Robot assisted gait training with active leg exoskeleton (ALEX). IEEE Transactions on Neural Systems and Rehabilitation Engineering 2009; 17(1): 2-8.##
  4. Kazerooni H, Steger L, Huang L. Hybrid control of the berkeley lower extremity exoskeleton (bleex), The International Journal of Robotics Research 2006; 25(6): 561-573.##
  5. Ghan, J, Steger R, Kazerooni H. Control and system identification for the Berkeley lower extremity exoskeleton (BLEEX). Advanced Robotics 2006; 20(9):  989-1014.##
  6. Aguirre-Ollinger G, Colgate JE, Peshkin MA, Goswami A. Active-Impedance Control of a Lower-Limb Assistive Exoskeleton.  IEEE 10th International Conference on Rehabilitation Robotics (ICORR); 2007. P. 188-195##
  7. Xiuxia, Y, Gui L, Zhiyong Y. Lower Extreme Carrying Exoskeleton Robot Adative Control Using Wavelet Neural Networks. Fourth International Conference on Natural Computation (ICNC); 2008. P. 158-167##
  8.  Cavallaro E.E, Rosen J, Perry J. Real-Time Myoprocessors for a Neural Controlled Powered Exoskeleton Arm. IEEE Transactions on Biomedical Engineering 2006; 53(11): 2387-2396.##
  9.  Kiguchi K, Tanaka T, Fukuda T. Neuro-fuzzy control of a robotic exoskeleton with EMG signals. IEEE Transactions on Fuzzy Systems 2004; vol. 12(4): 481-490.##
  10. George M. speed control of separately excited dc motor. Am J Appl Sci 2008; 5(3): 227-233.##
  11. Li F, Lau C.  development of fuzzy algorithms for servo systems. In: Philadelphia, pennsylvania. IEEE Int Conf Robot Autom; 1989. P. 65-71##
  12. Wolff K, Nordin P. Evolutionary learning from first principles of biped walking on a simulated humanoid robot. In: Chalmers University of Technology, Sweden. Proceedings of the business and Industry Symposium of the Simulation Technologies Conference; 2003. P. 31-36.##
  13. Changjiu Z, Qingchun M. Dynamic balance of a biped robot using fuzzy reinforcement learning agents. Fuzzy sets and systems archive 2003; 134(1): 169-187.##
  14. Manoonpong P, Geng T, Kulvicius T, Porr B, Florentin W. Adaptive, Fast Walking in a Biped robot under neuronal control and learning. PLos Computational Biology 2007; 3(9): 62-71. ##
  15.  Haavisto O, Hyotyniemi H. Simulation tool of biped walking robot model. [M.S.Thesis]. Finland: Helsinki University of Technology; 2004.##
  16. Katsuhiko Ogata. Modern Control Engineering. 4th ed. Prentic hall: academic press; 2010. p. 159-269.##
  17. Dzung Q, Phuong L. ANN Control System DC Motor. International Symposium on Electrical- Electronics Engineering ISEE 2005; 7(2): 25-34.##
  18. Khezri Sh, Anaraki H, Siddiqui A. Intelligent control of dc motor status. In: Conference on Electrical Engineering; University of Kerman. 2004. P. 16-25. [In Persian]##
  19. Horng H. Neural Adaptive Tracking Control of a DC Motor. Int J Comput Inform Sci Eng 1999;118(4):1-13. 65-71.##
  20. Roerdink M, Lamoth CJC, Kwakkel G, van Wieringen PCW, Beek PJ. Gait coordination after stroke. benefits of acoustically paced treadmill walking. Phys Ther 2007;87(3):1009–22.##
  21. Menhaj M. Computational Intelligence. Foundations of Neural Networks. 5th ed. Tehran: Amirkabir University Press; 1387. P. 153-200. [In Persian]##
  22. Nouri K, Dhaouadi R, Braiek N .Adaptive Control of a Nonlinier DC Motor Drive Using Recurrent Neural Networks. Applied Soft Computing 2008; 78(1): 371-382.##
  23. Atashpaz G. social optimization algorithm development and evaluation of its performance. [ MS Thesis]. Iran: faculty of Electrical and Computer Engineering Tehran University; 1387 [In Persian]##
  24. longo G. brescia M. multilayer perceptron. 2007. Available from: http://voneural.na.infn.it/mlp.html##
  25. Arumugam M, Rao C. On the improved performances of the particle swarm optimization algorithms with adaptive parameters, cross-over operators and root mean square (RMS) variants for computing optimal control of a class of hybrid systems. Applied Soft Computing Journal 2008; 8(1): 324–336.##
  26. Atashpaz-Gargari, C Lucas. Imperialist Competitive Algorithm: An algorithm for optimization inspired by imperialistic competition. IEEE Congress on Evolutionary Computation 2007; 10(4):  4661-4667.##
  27. Melkote H, Khorrami F. Nonlinear adaptive control of direct-drive brushless DC motors and applications to robotic manipulators Mechatronics. IEEE/ASME Transactions 1999; 2(1): 71-81. ##
  28. Pizzuit C. A Multi-objective Genetic Algorithm for Community Detection in Networks. In: U.S. state of New Jersey, Newark. 21st International Conference on Tools with Artificial Intelligence; 2009. P. 379-386.##
The Scientific Journal of Rehabilitation Medicine
Volume 4, Issue 1 - Serial Number 1
March and April 2015
Pages 49-62

Files

History

  • Receive Date: 20 July 2014
  • Revise Date: 12 September 2014
  • Accept Date: 09 October 2014
  • First Publish Date: 21 March 2015

Share

How to cite

Statistics