Time delays pose a significant challenge in control systems, especially in processes where delay dominates system behavior. Such delays can compromise system stability, cause oscillatory behavior, and degrade overall performance. This is particularly true of many chemical processes, where delay-dominant behavior is common and difficult to manage. Ensuring robust control in these scenarios is vital for reliable industrial operation. This study presents a novel dual-loop fractional-order control architecture based on a modified Smith Predictor. The conventional Smith Predictor is highly sensitive to time-delay mismatch and plant uncertainties, which degrade robustness and disturbance rejection. The proposed \(FO-TD\) (Fractional-Order-Tilt Derivative) approach employs fractional-order dynamics to enhance phase shaping and robustness, thereby maintaining stable performance under delay and model variations. The controller employs a dual \(FO-TD\) structure. The inner \(FO-TD\) component effectively compensates for delay, while the outer \(FO-TD\) manages the tracking and differential dynamics of the process. Controller gains are tuned using classical robustness criteria such as gain margin and maximum sensitivity, ensuring resilient performance under model uncertainties. The proposed design is validated through simulations on several benchmark chemical process models and further tested on a real-time experimental setup to demonstrate its practical viability. The proposed control system maintains robust performance even when plant parameters vary by as much as 40%. Further, the performance of the proposed controller is evaluated by comparing various performance error indices with those of existing designs.
