Digital Parameter-adaptive Control of Processes with Unknown Constant or Timevarying Dead Time

Abstract

Simultaneous estimation of process parameters and process time delay is regarded by applying a modification of the recursive least squares method. For an extended process model the order of the numerator polynomial is increased by an upper limit of the time delay. The estimation of the actual time delay d and the calculation of the final process model is done in every step by fitting the impulse response of a model with an assumed order mand time delay d to the impulse response of the estimated extended process model. The proposed method needs small computing time and small storage capacity .. By combining this method with a feedback control algorithm parameter-adaptive control algorithms result, which can be applied to processes with unknown parameters and unknown constant or timevarying time delays. The adaptive control of analog simulated processes with unknown constant and timevarying time delays in the presence of deterministic and stochastic disturbances is shown.