Robust adaptive PID control of functional electrical stimulation for drop-foot correction

A robust, adaptive proportional-integral-derivative (PID) control strategy is presented for controlling ankle movement using a functional electrical stimulation (FES) neuroprosthesis. The presented control strategy leverages the structurally simple PID controller. Moreover, the proposed PID controller automatically tunes its gains without requiring prior knowledge of the musculoskeletal system. Thus, in contrast to previously proposed control strategies for FES, the proposed controller does not necessitate time-consuming model identification for each patient. Additionally, the computational cost of the controller is minimized by linking the PID gains and updating only the common gain. As a result, a model-free, structurally simple, and computationally inexpensive controller is achieved, making it suitable for wearable FES-based neuroprostheses. A Lyapunov stability analysis proves uniformly ultimately bounded (UUB) tracking of the joint angle. Results from the simulated and experimental trials indicate that the proposed PID controller demonstrates high tracking accuracy and fast convergence.