Extremum-Seeking Control for a Robotic Leg Prosthesis with Sensory Feedback

By sensing changes in the contact force between the leg and level ground, humans can perceive their walking speed and adjust leg stiffness to accommodate walking terrains. To realize this natural regulation mechanism on the lower limb amputee, noninvasive functional electrical stimulation (nFES) was used to assist the subject in sensing the change in contact force between the leg and level ground, allowing for the adjustment of control parameters in the prosthetic leg. The cost function was designed to combine the tracking errors of the joints and changes in the stimulating current. For different walking terrains, an extremum-seeking control (ESC) method was employed to search for suitable control parameters in real time by monitoring the changes in the cost function. The stability of the proposed controller with extremum-seeking dynamics was demonstrated. The experimental results demonstrated that the extremum-seeking method effectively adjusted the control parameters of the prosthetic leg in response to changes in the cost function.