ADAPTIVE CONTROL OF CHEMICAL PROCESSES WITH VARIABLE DEAD TIME

New adaptive control and dead time compensation algorithms are developed for both single input--single output (SISO) and multiple input--multiple output (MIMO) processes with unknown or variable dead times. The SISO adaptive algorithm is based on a pole-zero placement controller/dead time compensator which is suitable for nonminimum-phase processes, requires no time-consuming controller design calculations, and includes a tuning parameter. On-line parameter estimation for the adaptive controller is performed with a recursive least squares algorithm capable of tracking sudden variations in the process parameters. The performance of the adaptive controller is verified by experimental application to a heat exchange control loop on a pilot scale distillation column. Extension of the SISO controller for MIMO processes yields the multivariable pole-zero placement controller/dead time compensator employed by the MIMO adaptive algorithm. The multivariable adaptive controller includes all of the features given above for the SISO algorithm, plus it allows for different time delays between the process inputs and outputs and it decouples the input-output pairs at steady state. A two input - two output distillation column simulation is employed to demonstrate the performance of the multivariable adaptive controller. Both the SISO and MIMO adaptive algorithms appear to be superior to algorithms previously developed.