M.Sc Thesis

M.Sc StudentGindin Igor
SubjectIntermittent Strategies for Human Motor Control:
Sensitivity to Model Uncertainties
DepartmentDepartment of Mechanical Engineering
Supervisor PROF. Miriam Zacksenhouse


The precision required to complete movements is not solely the purview of highly trained athletes. Instead, completing even the simple movements to pour a cup of coffee or even stand quietly require a multitude of sophisticated control mechanisms which are far from being precisely understood and modeled. Theories explaining human motor control are based on two different mechanisms: closed-loop versus open-loop control systems, with closed-loop theories are more prevalent. More specifically, the current most widely accepted theory of human motor control employs an optimal closed-loop system based on LQG with signal-dependent noise.

This research concentrates on a different control architecture - the intermittent control that combines open-loop and closed-loop control systems. Intermittent control was previously suggested as a possible model for human motor control since it explains phenomena that continuous control cannot explain (for example physiological refractory period). The performance of intermittent control in the presence of delays and plant uncertainties is evaluated using both analytical and numerical tools and compared with the performance of continuous control (in particular LQG). Analysis is limited to the simplified case of deterministic infinite time horizon that results in linear time-invariant (LTI) systems. Numerical simulations account also for the more complicated, finite time horizon case that results in time-variant systems. The trajectories that were produced in the finite horizon numerical simulations using the intermittent control are very similar to those presented in the literature (although modeled differently). These simulations indicate that probably intermittent control has a role in reaching movements strategy.