Optimal disturbance predictive and rejection control of a parabolic trough solar field

The control of field outlet temperature of parabolic trough solar field (PTSF) faces great challenges as the PTSF suffers from multiple field disturbances. To address the issue, this paper proposes an optimal disturbance predictive and rejection controller in a composite scheme, in which the field disturbances are firstly reconstructed into a single lumped one by model transformation, the lumped disturbance is then estimated by a third-order extended state observer with only the input-output data. Based on the current and past estimated lumped disturbance values, an incremental least-square support vector machine based predictor is designed to adaptively predict the future dynamics of the lumped disturbance. More importantly, an integration method of the estimated and predicted lumped disturbance sequence and optimal model predictive control is proposed based on real-time correction of steady-state sequence to compensate for the lumped disturbance feedforward while maintaining the optimality of the solution. The composite optimal control law demonstrates that the future dynamics of the lumped disturbance play an active role in disturbance rejection. The asymptotical practical stability analysis of the closed-loop system is presented rigorously. Some simulation cases are conducted on a verified analytical PTSF model to demonstrate the disturbance rejection and target tracking performance of our proposed design.

Automated summary

This paper proposes an optimal disturbance predictive and rejection controller in a composite scheme to address the control challenges faced by the parabolic trough solar field (PTSF). By reconstructing field disturbances, estimating and predicting lumped disturbance sequence, and real-time correction of steady-state sequence, the proposed design demonstrates effective disturbance rejection and target tracking performance.