Discrete event-driven control of an active orthosis regulated by electromyographic signals for Canis lupus familiaris

This study introduces the design of an asynchronous event-driven adaptive robust control that regulates a mobile limb orthosis position for the hind legs of a Canis lupus familiaris (CLF). The application of a suitable stability analysis based on a controlled Lyapunov function results in the laws to adjust the adaptive gains of a proportional integral derivative controller (PID). The controller succeeded in compensating external bounded perturbations and non-modeled uncertainties in the active orthosis device. This compensation forces the tracking between the current positions and some reference trajectories obtained by a biomechanical gait cycle analysis of the CLF. The controller starts with the event triggered from the power of the electromyography signal from the frontal legs. If the signal power is higher than a predefined threshold, the movement of the orthosis will initiate. Electromyographic signals were acquired offline and injected into a virtualized orthosis model to test the event-driven control design. A set of numerical simulations confirmed a better performance of tracking reference trajectories and the effect of the event-driven controller on the orthosis operation. The experimental validation of the proposed output feedback controller on the designed orthosis seems to justify a potential automatized rehabilitation therapy based on the proposed electromyography-driven strategy.