Model-Free Control of Finger Dynamics in Prosthetic Hand Myoelectric-based Control Systems

This paper proposes an approach to the tuning of model-free controllers for the midcarpal joint angles, which arc important finger angles that ensure the desired finger dynamics in prosthetic hand myoelectric-based control systems. The process in these control systems is characterized by fuzzy models that operate with myoelectric signals obtained from eight myoelectric sensors and past inputs and/or outputs, where the controlled outputs arc five finger angles. Since the fuzzy models exhibit very good performance as shown in authors' recent papers that produced evolving fuzzy models, they are used to simulate the process behaviour. The Multi Input-Multi Output (MIMO) control system structure consists of five separate Single Input-Single Output control loops with the most simple model-free controllers represented by intelligent Proportional (iP) controllers, separately designed and tuned for each finger. Digital simulation results are included to suggestively illustrate the very good perfonnance of the control systems with iP controllers. The MEMO control system performance is compared with that of the same system but with Proportional-Integral controllers, which are optimally tuned in a model-based manner by a metaheuristic Grey Wolf Optimizer (GWO) algorithm. The fair comparison is ensured by the optimal tuning of the free parameters of iP controllers in a model-based manner using GWO.