Predictive controller for temperature regulation of a solar receiver

Solar energy is the largest renewable resource freely available for exploitation. Solar thermochemical processes have the potential to transform solar energy into storable and transportable fuels. Unfortunately, this solar thermal technology is faced with one critical challenge which relates to the intermittency of solar radiation that adversely affect temperature stability of the solar receiver. Fortunately, application of variable aperture mechanism to regulate the light entry into solar receiver has emerged as a promising solution to this drawback. Additionally, performance of such a controller heavily depends on the characteristics of the light source and radiation properties of the cavity receiver. Therefore, in order to show the versatility and capability of this technique, it has to be examined using the most commonly practiced solar thermal systems, which include: a solar furnace housing a parabolic dish and a high flux solar simulator.

Nesrin Ozalp, Professor of Mechanical and Industrial Engineering at University of Minnesota Duluth in collaboration with Dr. Hamed Abedini Najafabadi at University of Leuven in Belgium developed a predictive controller for temperature regulation of a solar receiver. The researchers hoped that the controller would compensate the effects of fluctuation in incoming solar energy from sunrise to sunset by changing the receiver aperture size. To this note, they utilized facilities of different radiation sources, total power input, reflector type and flux distribution. Their work is now published in the research journal, Solar Energy.

The research team used Monte Carlo ray tracing technique for optical analysis of both radiative heat sources. Gaussian curve was then used to approximate the concentration ratio at the focal plane of the parabolic dish for the Paul Scherrer Institute solar dish. A comparison of simulated and experimentally measured heat flux distribution confirmed the accuracy of the optical analysis done by Monte Carlo ray tracing method. They then developed a numerical model which was validated by conducting several experiments with varied aperture sizes. The controller was then constructed based on nonlinear Hammerstein model where its parameters were identified according to the results obtained from dynamic numerical model. Eventually, the performance of this adaptive predictive controller was compared with a regular proportional integral derivative (PID) controller for both radiative heat sources.

From the simulation results, the authors observed that in a special case where the proportional integral derivative controller failed to regulate the temperature due to the nonlinear behavior of the system, accurate control was achieved by adaptive nonlinear model predictive controller. They also noted that, integral absolute error performance index, settling time and overshoot of the predictive controller were lower than the proportional integral derivative controller which indicate that adaptive nonlinear model predictive controller performs better. Moreover, simulations for an actual sunny day confirmed that the adaptive predictive strategy exhibits reasonable performance during start up, shut down and normal operation periods. Furthermore, simulation for a partially cloudy day indicated that the control system was able to compensate high fluctuations in solar radiation and keep the temperature very close to the set-point.

Ozalp and Abedini Najafabadi study successfully presented a cross examination of an adaptive model predictive controller for temperature regulation of a solar receiver. The controller exhibited robust performance during sunrise and sunset times as well as passing clouds conditions (when fluctuations in solar radiation are expected).