Towards an improved gain scheduling predictive control strategy for a solar thermal power plant

Abstract

This study improves a recently proposed gain scheduling predictive control strategy for the ACUREX solar collector field at the Plataforma Solar de Almería, Spain. Measured disturbances are an integral part of the plant and while simple classical, series and parallel, feedforward approaches have been proposed and used extensively in the literature, the proposed approach incorporates a feedforward systematically into the predictive control strategy by including the effects of the measured disturbances into the predictions of future outputs. Models of the measured disturbances are estimated around a family of operating points directly from input–output data and using a subspace identification method while taking into account the frequency response of the plant. Input–output data are obtained from a validated non‐linear simulation model of the plant. The non‐linear simulation model is validated against measured data obtained from the ACUREX plant and the effectiveness of the proposed control approach is evaluated in the same non‐linear simulation environment. The study also considers related issues such as the significance of sufficient modelling of the measured disturbances and the impact of incorporating the expected future behaviour of a measured disturbance along a given prediction horizon, a theme which has received little attention in the literature.