Optimization Parameters of PID Controller for Powered Ankle-foot Prosthesis Based on CMA Evolution Strategy

Optimization parameters of PID controller based on Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is presented in this paper. It is used to solve the problem of torque control for powered ankle-foot prosthesis. Original optimization parameters method of PID controller for powered ankle-foot is time-consuming and cannot get satisfied control effect. The parameters of PID control are used as an individual of CMA-ES in this paper. Appropriate fitness function is selected to adjust the PID parameters online. Step signal and torque approximation are used as the system input to verify the controller performance. In unit-step response, the overshoot of original PID is 15 times as much as it of CMA-ES PID, the setting time of original PID is 6 times as much as it of CMA-ES PID. In device torque response, the output of CMA-ES PID is stabilized throughout the control process. These indicates that CMA-ES PID is an effective control strategy for torque control of powered ankle-foot prosthesis.

[1]  Samuel Kwok Wai Au,et al.  Powered ankle-foot prosthesis for the improvement of amputee walking economy , 2007 .

[2]  Edwin van Asseldonk,et al.  Improving the transparency of a rehabilitation robot by exploiting the cyclic behaviour of walking , 2013, 2013 IEEE 13th International Conference on Rehabilitation Robotics (ICORR).

[3]  Michael Lars Palmer,et al.  Sagittal plane characterization of normal human ankle function across a range of walking gait speeds , 2002 .

[4]  Nikolaus Hansen,et al.  Adapting arbitrary normal mutation distributions in evolution strategies: the covariance matrix adaptation , 1996, Proceedings of IEEE International Conference on Evolutionary Computation.

[5]  Mohammad Pourmahmood Aghababa,et al.  Optimal design of fractional-order PID controller for five bar linkage robot using a new particle swarm optimization algorithm , 2015, Soft Computing.

[6]  Petros Koumoutsakos,et al.  Reducing the Time Complexity of the Derandomized Evolution Strategy with Covariance Matrix Adaptation (CMA-ES) , 2003, Evolutionary Computation.

[7]  Michael Goldfarb,et al.  Realizing the Promise of Robotic Leg Prostheses , 2013, Science Translational Medicine.

[8]  Nikolaus Hansen,et al.  Completely Derandomized Self-Adaptation in Evolution Strategies , 2001, Evolutionary Computation.

[9]  Paolo Fiorini,et al.  Robust Force Control of Series Elastic Actuators , 2014 .

[10]  R. Waters,et al.  The energy expenditure of normal and pathologic gait. , 1999, Gait & posture.

[11]  Rachel W Jackson,et al.  Human-in-the-loop optimization of exoskeleton assistance during walking , 2017, Science.

[12]  H. van der Kooij,et al.  Reference Trajectory Generation for Rehabilitation Robots: Complementary Limb Motion Estimation , 2009, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[13]  Hartmut Geyer,et al.  A Muscle-Reflex Model That Encodes Principles of Legged Mechanics Produces Human Walking Dynamics and Muscle Activities , 2010, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[14]  Joshua M. Caputo,et al.  Increasing ankle push-off work with a powered prosthesis does not necessarily reduce metabolic rate for transtibial amputees. , 2016, Journal of biomechanics.

[15]  N. Yadaiah,et al.  An optimized relation between Ti and Td in Modified Ziegler Nichols PID controller tuning , 2013, 2013 IEEE International Conference on Control Applications (CCA).

[16]  Steven H Collins,et al.  A universal ankle-foot prosthesis emulator for human locomotion experiments. , 2014, Journal of biomechanical engineering.

[17]  Daniel P Ferris,et al.  An improved powered ankle-foot orthosis using proportional myoelectric control. , 2006, Gait & posture.

[18]  Christopher L. Vaughan,et al.  Dynamics of human gait , 1992 .