Finite state control of FES systems: application of AI inductive learning techniques

A technique is described for inductively deriving decision rules for gait event detection in an FES (functional electrical stimulation) control system, based on a collection of examples of the events to be detected. Results are presented for learning the control strategy of a manually operated two-channel stimulation system, and the general applicability of this technique to the control of FES orthoses is outlined. It is shown that inductive learning techniques represent an efficient and accurate method of deriving rules for the control of FES. In the example described the knowledge gained by the subject´s experience of controlling the stimulator is effectively captured by the program