Comparison with the PID Controller

The following values have been adopted for the PID gains gp = 10, g = 5 10-1, and go = 10-3. In order to perform a fair comparison between the PID approach and the two-loop controller-observer strategies, the PID gains have been tuned via a trial-and-error procedure so as to achieve the same control effort of the other considered control strategies (i.e., so as to obtain the same time histories of u as far as possible). 

The results are shown in Fig. 5.9 as a function of the set-point temperature for both the model-based and the PID controller. In both cases, the critical 6 increases with the setpoint temperature since a more efficient heat exchange is required when 

In this chapter an overview of the most widely adopted temperature control schemes for chemical batch reactors has been provided. Moreover, an adaptive model-based controller-observer approach has been proposed, analyzed, and compared to other approaches. 

The approach is developed for a fairly wide class of processes, i.e., the class of irreversible nonchain reactions characterized by first-order kinetics. Although this is not the most general case, it encompasses several real reactive processes. 

Moreover, it is worth remarking that the use of an accurate and reliable state observer, which is necessary for the proposed controller, can be advantageous for other purposes as well (e.g., process monitoring and fault diagnosis). 

---