COMPARISON OF H-INFINITY AND ENSEMBLE KALMAN FILTER FOR ESTIMATING MOTION OF MIDDLE FINGER

Abstract

Upper extremity paresis is a condition in which a person experiences muscle weakness in one or both hands. This condition can cause impairment in motor function, hinder daily activities, and affect the life quality of the sufferer. In some cases, paresis can result from nerve injury, neurological disease, or an accident. To help improve the life quality of the sufferer experiencing upper extremity paresis, the development of the Finger Prosthetic Arm Robot, an assistive robotic hand designed to provide assistance in the movement of the finger experiencing paresis, is required. This technology aims to restore its functional ability and the independence of the patient in performing daily activities, such as picking up objects, grasping, and performing other precise movements. The main purpose of this paper, the researcher compared two methods to estimate the motion of the middle finger robot, that is, the H-infinity method and the Ensemble Kalman Filter (EnKF) method. The simulation results show that both methods had almost the same accuracy, and the simulation by generating 800 ensembles was more accurate than that by generating 400 ensembles with an accuracy difference of about 10% above the accuracy rate of 98%.