Reconstruction of human neuromuscular control signals using a sliding mode control technique

A sliding mode controller, developed using MATLAB/SIMULINK, is incorporated within an existing model of the complex nonlinear system of the human leg with the aim of identifying the operation of the feedback control system within the body to produce the observed leg movement during walking. The control design procedure first defines a switching function, the so-called sliding surface, which prescribes the required dynamic behaviour of the error between the measured data and the model output. A discontinuous feedback control law is then designed to drive the system trajectories onto the sliding surface and force them to remain there. The attractiveness of sliding mode control for the analysis of biomechanical control systems arises from the equivalent control principle. If a switching surface is selected which ensures exact tracking between the model trajectory and experimentally measured trajectory from a human subject, then the filtered discontinuous control signals from the simulation model provide estimates of the signals required to achieve the observed movement. The inherently robust properties of sliding mode control ensure precise tracking by the model of the actual leg movement, thus enabling the feedback control signal required to carry out the movement to be reconstructed.