*4.6. Remarks on Comparative Simulation Results*

The simulations results and closed-loop performance analysis shows that some of these autotuning methods allow for greater flexibility in the design, such as [34,47,49]. A faster settling time is obtained in all case studies using the autotuning method in [40]. The drawback consists in a larger control effort. The simple tuning rules from [32,42] are generally outperformed by the other autotuning methods reviewed, except for delay dominant systems, where the performance is close to the best one. For higher order systems and poorly damped ones, the best closed-loop results are obtained using the autotuning methods in [36,49] for both reference tracking and disturbance rejection. For FOPDT delay-dominant processes, the results show that the parameters of the fractional-order controllers should be estimated using the autotuning methods in [26,36,49]. In this case, improved reference tracking and disturbance rejection are obtained. For integrating timedelay processes, the best results in terms of overshoot are obtained using either a FO-PID determined based on the autotuning method in [47] or in [49]. The best settling time is obtained using either the autotuning method in [40,47]. However, the FO-PID controller autotuned according to [47] requires a significant control effort, larger than those in [40,49]. For FOPDT lag-dominant processes, the autotuning methods from [26,27,49] ensure a small

overshoot, whereas fast settling and disturbance rejection times are achieved using FO-PIDs determined according to [26,27,40]. However, the control effort required when using the FO-PIDs tuned based on [27,40] is larger compared to the FO-PID tuned using the approach in [26].
