A novel unified method for time-varying dead-time compensation

This paper proposes an unified procedure for time-varying dead-time compensation. The method is an adapted version of the Filtered Smith Predictor (FSP), which is coupled to a delay estimation scheme and two adaptive Linear Parameter Varying (LPV) blocks. These LPV blocks allow the DTC to autonomously regulate the amount of closed-loop robustness, with respect to the (estimated) amount of delay. The method is easily adjustable and can be tuned to provide a balance between robustness and performance objectives, while being able to deal with unstable and integrative processes. A series of numerical simulations are included to illustrate the advantages of the proposed method towards reference tracking, noise attenuation, disturbance rejection and uncertainty handling. This novel DTC enables enhanced performances with respect other methods from the literature.