Heliostats of central receiver solar power plants reach dimensions up to 150 m2 with focal lengths up to 1000 m. Their optical properties and tracking accuracy have great influence on the power plant efficiency and need to be monitored both at plant start up and during operation. Up to now, there are few efficient and fast measurement techniques that allow the heliostat properties to be measured. Flux density measurements and close-range photogrammetry are possible approaches, yet they do not fulfill the requirement to be accurate, inexpensive and fast at the same time. In this paper, we present a non-contact measurement principle, which uses edge detection to extract the heliostat and facet vertices. This information is used to calculate the surface normals. Furthermore, the corners can replace retroreflective targets generally used in close-range photogrammetry, thus enabling a fast and completely automatic evaluation of the three-dimensional heliostat structure. The pictures are provided by a digital camera which is mounted on a pan tilt head on top of the central receiver tower, offering visibility to all heliostats and allowing the automated qualification of whole heliostat fields in a short period of time. It is shown that measurement uncertainties in heliostat orientation for the investigated heliostat are below 4 mrad in 80% of the relevant heliostat positions. Heliostat orientation is available within three minutes. Photogrammetric measurements based on edge detection at a 40-m2 CESA-1 heliostat at the Plataforma Solar de Almeria (PSA) exhibit an accuracy of 1.6 mrad for a single-facet normal vector with the results being available within 30 minutes. The reduced measurement time allows the economic characterization of entire heliostat fields. The lower accuracy compared to manual photogrammetry with retroreflective targets is still sufficient to detect facet misalignments in existing heliostat fields.
[1]
Eckhard Lüpfert,et al.
Photogrammetry: A Powerful Tool for Geometric Analysis of Solar Concentrators and Their Components
,
2005
.
[2]
Jeffrey C. Lagarias,et al.
Convergence Properties of the Nelder-Mead Simplex Method in Low Dimensions
,
1998,
SIAM J. Optim..
[3]
Christopher R. Houck,et al.
A Genetic Algorithm for Function Optimization: A Matlab Implementation
,
2001
.
[4]
Steffen Ulmer,et al.
Automatisierte hochaufgelöste Vermessung der Spiegelfehler von Heliostaten
,
2007
.
[5]
Rafael C. González,et al.
Digital image processing using MATLAB
,
2006
.