An LMI-based disturbance reduction control for systems with multiple state and input delays

A disturbance reduction control technique is proposed for systems with multiple states and input time delays to minimize the adverse effects of unmeasurable deterministic disturbances of known waveforms but unknown magnitudes and arrival times. This scheme is based on reconstructing the disturbance state through a multiple time-delay observer, which is designed using a linear matrix inequality (LMI) approach and using part of the control signal to actively minimize the Euclidean norm of the disturbance vector estimate.