Iterative Learning Control as an Intervention Aid to Stroke Rehabilitation

This paper describes the experimental test facility that has been designed as part of a current project whose aim is to investigate the use of iterative learning control (ILC) and related strategies to mediate the electrical stimulation applied to a number of muscles of the shoulder and upper limb of stroke patients. ILC is a technique that has its origins in the area of industrial robotics and is especially targeted at systems operating in a repetitive mode with the additional requirement that a specified output trajectory over a finite interval (or trial) is followed to a specified level of precision. Motivated by human learning, the basic idea of ILC is to use information from previous executions of the task in order to improve performance from trial to trial. In the context of this work, ILC offers the opportunity of allowing the user to learn how to perfect a given task by controlling the amount of additional stimulation applied from one trial to the next. If this has been achieved, then further refinement is possible to reduce the stimulation effort supplied by the control scheme and thereby increase the effort necessary from the subject. In particular, ILC requires a basic representation of the underlying (dynamic) relationship between stimulation (or input) and response (or output) and also the specification of a target (or reference) trajectory to form the goal for the ILC scheme to achieve