Session: 18-03 HelioCon Solar Field

Paper Number: 130671

130671 - Heliocon Closed Loop Control: Extremum Seeking Control Small-Scale and Single Heliostat Testing 

Abstract: 

In this paper, we will overview the development and implementation of an Extremum Seeking Control (ESC) algorithm used for closed-loop control of a heliostat. This paper will include an overview of the algorithm’s development, the challenges and implementation into a LabView framework, design of an experimental setup, and experimental results using a small-scale testbed and a single heliostat for tracking. ESC is an optimization algorithm that seeks to maximize an unknown power function. The input into the algorithm is an initial azimuth and elevation and the output is an azimuth and elevation with more power. In the case of the NSTTF Solar Tower, the power distribution can take the form of measurements of pixel intensity, heat flux, or temperature. Since the power function is non-explicit, it isn’t feasible to take the derivative to find the gradient, therefore we must use a gradient estimation algorithm utilizing the least-squares approach to estimate using sampled data points of power from a solar beam on the Solar Tower. ESC will then use the estimated gradient and the heliostat dynamics to adjust the azimuth and elevation of the heliostat to move the solar beam to the optimal position that maximizes the power output. The least-squares estimation algorithm has been validated using a multivariate gaussian distribution developed in MATLAB. The simulation began at a chosen pixel location (x, y) away from the optimal. The ESC algorithm then used the gradient estimation to seek and converge to the optimal power location. Two experimental setups will be used for validation of the algorithm. The first is a small-scale experiment designed to emulate a stagnant sun and dc motors on tracks to move a BCS camera to the location of the maximum light intensity. The second experimental setup will be a single heliostat tracking the sun to keep the maximum power received on the tower at the location desired. These two experiments will be used to validate the ESC algorithm and give confidence for future work scaling this algorithm to be used in multiple heliostats.

Presenting Author: Kenneth Armijo Sandia National Laboratories

Presenting Author Biography: Dr. Kenneth Armijo is a systems engineering staff member who leads molten salt and molten alkali metals R&D at the National Solar Thermal Test Facility (NSTTF). His research interests are in alternative energy technologies and sustainability, as they pertain to scientific and technological innovation, business and policy. Dr. Armijo holds a Ph.D. in Mechanical Engineering from the University of California, Berkeley with minors in Energy and Resources, and business credentials in Management of Technology (MOT) from Berkeley's Haas School of Business. Dr. Armijo’s research in concentrating solar power (CSP) and nuclear energy (NE) consists of system design for high-temperature (>720 °C) thermodynamic and commercial R&D systems, employing nitrate, chloride and fluoride molten salts and alkali metals (sodium) as the heat transfer fluid. He currently is the PI for multiple U.S. DOE projects, including the DOE Advanced Salt Valve project, which is a collaboration between CSP and Nuclear Energy industrial partners. His research has also consisted of falling particles for centralized concentrating solar receivers. He also leads research activities pertaining to solar Stirling Engine applications as well as for solar reactor R&D and high-flux materials characterization. Dr. Armijo also serves as a lead Test Director for high-temperature materials research for power towers and solar furnace investigations.
In addition to CSP, his research also includes, and has included Photovoltaics (PV) and Distributed Energy Technologies investigations pertaining to spectral derates phenomena of single, and multijunction devices, arc-fault plasma reliability physics, thermal phenomena of PV technologies and inverter/power electronics reliability research. His arc-fault analytical and experimental work also spans applications in Nuclear energy as well as other high-voltage/high-current DC power reliability applications, where he is recognized for work in the development of U.S. and international arc-fault detection and mitigation codes and standards. His CSP and PV work addresses his vision for a more economic and sustainable future, both locally and globally.

Authors:

Kenneth Armijo Sandia National Laboratories
Haden Harper Sandia National Laboratories
Zachary Bernius Sandia National Laboratories
Luis Garcia-Maldonado Sandia National Laboratories
Ansel Blumenthal Sandia National Laboratories
Aaron Overacker Sandia National Laboratories
Anthony Evans Sandia National Laboratories
Claus Danielson The University of New Mexico