Data-Driven Controller Synthesis for Parameters Unknown Linear-delay Systems with Input Constraint

In practical control engineering, delay and actuator saturation are main challenges which may lead to system performance degradation, or even controlled variables divergence. Combination of Smith predictor and model recovery anti windup method is an effective solution to linear delay system anti-windup synthesis, however model dependent property of the above two methods brings obstacles for the application and promotion. To overcome such a difficulty, a data-driven approach is proposed in this paper, operating data is collected and utilized to simultaneously tune feedback controller parameters and estimate internal model applied in Smith predictor. Subsequently, anti-windup compensator is constructed with the estimated model and a compensator gain optimization structure is proposed based on Lyapunov equation. Finally, simulations have been implemented to three typical processes in chemical engineering, and detailed performance comparison is provided to illustrate the effectiveness of the proposed method.