Robust Fractional Order PID Controller Coupled with a Nonlinear Filtered Smith Predictor for Solar Collector Fields

This work introduces a Fractional Order PID (FOPID) controller associated with a nonlinear Filtered Smith Predictor (FSP) to control a Solar Collector Field (SCF) system. The combined approach enhances the control performance by decoupling model uncertainties. First, the FOPID addresses linear model uncertainties through optimal robust design across the operational conditions, while the FSP compensator adeptly manages the system's varying time delays. Rigorous tuning in the frequency domain considering closed-loop, robustness, and disturbance rejection performances is developed using a bi-level optimization approach to obtain the FOPID control parameters. Experimental results of the FOPID + nonlinear FSP demonstrate low rise time for reference tracking and robust stability across several operating conditions, even in the face of high-frequency disturbances in the solar irradiance due to passing clouds. Moreover, the nonlinear FSP effectively handles variations in time delays undergone by the water flow variation during the experiment. The achieved outcomes of the overall control structure, comprising the FOPID FSP, demonstrate a promising approach for SCF system control, obtaining robust behavior in the presence of model uncertainties and disturbances while offering satisfactory reference tracking capabilities across an extensive range of plant operations.