Application of a model of cerebellar function to the maintenance of human upright posture

Abstract

In this thesis a simple human postural control model is suggested and analyzed based on hypothesized neurophysiology of the cerebellar function and the musculoskeletal system. The cerebellum model is made up of simple linear filters such as differentiator and integrator. The simple linear filters implement a linear feedback control scheme including a phase lead compensator. The neural feedback signal represents the action of the cerebellum in the processing of angular position and angular velocity error signals. The goal of the investigation is to indicate whether the simple linear filters can describe neurophysiological functions of the cerebellum to compensate for the neural delays and coordinate the postural strategies that make possible human upright posture in gravity. Performance of the model is investigated with regard to disturbance rejection after adjustment of the parameters representing the cerebellum and the muscle. Whether the combination of the cerebellar and musculoskeletal control systems can realistically model human posture balance recovery is evaluated by simulating human postural maintenance during backward translation of a support surface. The simulation is compared with actual human postures and movements. The simulation realizes the ankle and hip strategy that prevails in human posture, and suggests the functions of the cerebellum.