Recursive estimation for system of unknown delay
Ashraf Y. Elnaggar, Guy Albert Dumont, Abdel Latif Elshafei · 2003
An algorithm is presented for the recursive estimation of both process parameters and delay when no a priori knowledge of the delay is required. The algorithm is a modification of any standard recursive parameter estimation algorithm and can be applied easily to any of the well-established versions of estimation algorithms (least square, instrumental variables, maximum likelihood, etc.) and their extensions. The algorithm requires minimum additional computations compared to the original algorithms. The delay is explicity estimated, and the parameter vector is of the same length as for the undelayed system. This means clear separation between the system dynamics and delay. Simulation results showing the good performance of the algorithm are presented.>