Experimental evaluation of predictive temperature control for a batch reactor system

A nonlinear generalized predictive control (NLGPC) algorithm developed by Katende and Jutan (1996) is tested on a nonlinear batch reactor system by carrying out a number of experiments and comparing its performance with other control strategies. This experimental setting allows a more practical comparison of the relative merits of these controllers. The NLGPC is shown to outperform the constrained self-tuning PID (STPID) controller by Katende and Jutan (1993), and the generalized minimum variance (GMV) controller by Clarke and Gawthrope (1975). It is also shown to have better performance than the well-known generalized predictive control (GPC) algorithm by Clarke et al. (1987). The advantage of the NLGPC over the other controllers is attributed to its adaptive nature and use of nonlinear process models in its design.