Work

Inlet-Temperature Disturbance Rejection in Solar Collector Fields Using Learning-Based Practical Nonlinear Predictive Control

Year: 2026
Type: article
Abstract: Rapid variations in the Heat-Transfer-Fluid (HTF) inlet temperature can disrupt outlet-temperature regulation in solar collector fields, particularly when fast recovery interacts with delayed control action. This work assesses the previously developed Learning-Based Practical Nonlinear Model Predictive Controller (LBPNMPC) for rejecting such disturbances, motivated by expansion-vessel fluid-exchange transients observed at an actual facility. A matched clear-day simulation uses fourteen inlet-temperature disturbances generated from five experimentally measured profiles. LBPNMPC is compared with conventional filtered-error PNMPC, both with measured-disturbance feedforward, and with PNMPC without disturbance feedforward. In this matched simulation, LBPNMPC reduces the inlet-event integral absolute error by 34.3%, integral squared error by 49.3%, mean recovery time by 23.9%, and HTF-flow total variation by 11.4% relative to conventional filtered-error PNMPC. Experimental operation is assessed at the CIESOL research centre, University of Almería, Spain. During a severe 9.27 °C inlet-temperature decrease, the corresponding measured outlet-temperature decrease was 5.60 °C, approximately 60% of the inlet-temperature variation. The experiment demonstrates the practical applicability of LBPNMPC during a severe inlet-temperature disturbance, complementing the quantitative comparison with conventional PNMPC provided by the matched simulation.
Source: Energies
Language: English
Field: Energy
Open Access status: gold
Estimated APC paid: $3,137
OpenAlex ID: W7236709706
FWCI:
Cites: 25
Cited by:
Related to: 10