Neuro-Dynamic Control of an Above Knee Prosthetic Leg

The control of a prosthetic leg for above-knee amputees is addressed in this paper. Simultaneous control of the knee and ankle joints is difficult because the desired gait of the user is unknown and changes with user intent. Further, the dynamics between the knee, ankle, and the upper body are coupled and unknown. In this paper, we propose a novel neuro-dynamic control strategy that can guarantee stable control of the prosthetic limb while minimizing the gait asymmetry between the intact and prosthetic limb. Using an actor-critic network, the method in this paper learns the required control actions while minimizing a long-term cost that is indicative of the gait asymmetry. The stability of the prosthetic system and its performance are rigorously analyzed and the efficacy of the proposed approach is demonstrated through numerical simulations. It is anticipated that minimizing the asymmetry in gait between intact and prosthetic side will improve user safety, lower metabolic energy consumption, and reduce musculo-skeletal pain. In the long term, this will reduce instances of cardio-vascular and other health complications.