Search

Solar field performance is dependent upon the ability of each heliostat to project a concentrated beam shape on the solar receiver. The specified tolerances for the heliostat surface are very tight since a deviation of even tenths of a milliradian is significant for performance. A precise method of measuring the actual heliostat surface is critical both for quality assurance and in order to calculate the actual flux that the solar field will apply to the solar receiver. The measurements will be used as part of the solar field control so that the flux can be distributed as required for solar steam generation. BrightSource has developed a methodfor estimating the heliostat's shape after installation in the solar field. The measurement system is based on a heliostat control system, a visual range camera, the heliostat itself and the sun. The camera captures a rapid sequence of images as the heliostat moves in a precisely-defined path. In parallel, the system captures the exact time of each image. Throughout the sequence, the system also records the movement of the heliostat in both elevation and azimuth directions. The captured images, the timestamps and the matching recorded movements are post-processed using a complex algorithm to calculate the surface of the heliostat mirrors. The algorithm divides the mirror surface into many different elements. The algorithm identifies each element in each of the sequence of images and uses the recorded data to calculate the normal for that element. On the basis of all the elements, the algorithm generates a unified mirror surface. This methodcan be used both as a quality control method (sample of solar field after construction) and as a method of calculating the baseline characteristics of each heliostat.

Dissertation, RWTH Aachen University, 2017; Dissertation, RWTH Aachen University, 2017, Solar thermal power plants employ sunlight directly to generate electricity. Although the solar energy is free of charge, the plant needs to operate cost-efficiently in order to compete with conventional coal-, gas-, and nuclear energy. One possibility is the optimization of the components used. Therefore, minimizing thermal and optical losses is essential. Different loads and boundary conditions have a significant influence on shape accuracy of parabolic solar mirrors. In general, it depends substantially on the support structure which usually differs from those used in standardized laboratory measurements. The aim of this work is to characterize influences on the shape accuracy in collector, the preparation of appropriate models and programs, as well as the evaluation of how well mirror shape in collector can be predicted based on laboratory results. The optical performance of the mirrors as well as the annual yield are quantified.As a result of this work, interfaces to the DLR internally developed tools SPRAY and Greenius are created and comprehensive workflows employing Matlab tools provided. In laboratory and collector test benches optical measurements of the mirror shape, the support structure and occurring strains at the glass brackets holding the mirrors are carried out and reproduced with FE-models. An optimization workflow is implemented to support the evaluations. The investigations were carried out with the EuroTrough-geometry by using mirrors of RP 3 geometry.In the following, important results of this work are summarized: As an important finding, the consideration of further parts of the support structure than the glass brackets does not significantly influence dead load results. However, and as presented in exemplary simulations, geometrical deviations of the mounting elements of the support structure can have a significant effect on the shape accuracy. The resulting optical performance as well as the influence on the annual yield for the investigated load cases is evaluated. Mounting inaccuracies of the mirror mounting elements can induce significant shape deformations even in case of only a few mrad or mm deviation - and also affect the optical performance and the annual yield of the plant. Deriving technical specifications, but also the optimization of the geometries is made possible, since optical, thermal, and economical losses can be quantified due to the progress within this work. The predictability of mirror shape in collector is limited, and will be discussed, since the aforementioned geometrical deviations of the support structure are not universally known or predictable.

Deviations from the ideal shape of reflector panels for parabolic trough solar power plants have relevant impact on field efficiency and thus on the performance of the whole power plant. Analyzing the gravity-induced deformation of mirror shape for different mirror angles is relevant for performance calculation of solar parabolic trough collectors and identifying optimization potential of the mirror panels. Two mirror model versions (stiff and elastic supports) are evaluated in four angles: in horizontal laboratory angle (mirrors facing upward with mounting points horizontally aligned), and in 0°, 45° and 90° collector angle. The resulting slope maps are calculated in a separate post-processing. In order to evaluate the effect of gravity load on mirror shape, the deformed mirror in each evaluated angle is compared to the non-deformed mirror shape, and to the shapes in 0° (zenith) collector angle, respectively. The resulting slope deviation maps show the mirror deformation in different mirror angles. Stiffness of the mounting to the support structure has a relevant impact. Mirror deformation on elastic brackets (SDx up to 1.6 mrad) is much more pronounced than on an ideal stiff support structure (SDx up to 1.0 mrad).

The present thesis examines the influence of measurement parameters and boundary conditions on shape accuracy of parabolic trough concentrator mirrors of RP3 geometry by means of optical measurements and finite element analyses. Deflectometric measurements of mirror panels are performed in common laboratory setups: in vertical and horizontal measurement position, both with and without tightening the mirrors to a support frame with screws. The results serve for validation of finite element models which are applied to calculate the difference in shape accuracy between, firstly, the individual setups and secondly, between mirrors horizontally fixed to differently rigid support frames. The effect of angular deviations of the mounting pads is assessed. Gravity-induced deformation of mirrors mounted to two exemplary collector support structures is determined for the relevant working positions. The transferability of shape accuracy data measured in laboratory to data applying for mirrors mounted to a collector is assessed. For RP3 inner mirror panels a difference of up to 0.8 mrad in root mean square slope deviation (SDx) and 3.3 mm in root mean square focus deviation (FDx) from vertical to horizontal position is measured. Gravity-induced deformation and resulting slope and focus deviation on three different types of horizontally oriented support frames is modeled: an ideally rigid support frame, a laboratory support frame, and an ideal support frame with elastic brackets. A variation of 0.5 mrad (SDx) and 1.8 mm (FDx) is calculated for perfectly shaped RP3 inner mirrors mounted onto an ideally rigid support frame compared to the case when mounted to a support frame with elastic brackets. In case of an angular deviation of the mounting pads mirror panels fixed to an ideally rigid support frame show additional local slope deviation in the magnitude of the angular pad deviation. Due to gravity-induced deformation and slope deviation the shape accuracy of RP3 inner mirrors mounted to an ideally rigid collector support structure varies up to 1.0 mrad in SDx and up to 1.2 mrad in SDx when compared to the ideal shape and the shape in zenith collector angle, respectively. For RP3 inner mirrors mounted to a collector structure with elastic brackets the values are 1.6 mrad and 2.1 mrad. Whereas the models of mirrors mounted to a laboratory support frame allow for a conversion and thus a comparison of results obtained in different setups, the models for the collector support structure need to be refined in order to allow for a conversion from laboratory to collector measurement results. The results demonstrate that measurement position, mounting mode and employed support structure have to be documented in order to obtain reproducible measurement results. They should hence be included as measurement boundary conditions in guidelines for the measurement and assessment of mirror shape accuracy for concentrating solar collectors. The inspection of the angular accuracy of the mounting pads is part of a complete optical mirror assessment. The data of standardized shape accuracy measurement serve for subsequent calculation of intercept factors and annual energy yields as well as for the evaluation of optimization approaches.

Next generation telescopes are to employ segmented mirrors to realize extremely large primary mirror surfaces. For example, one of the current next generation large telescope projects, the Thirty Meter Telescope (TMT), will employ a 30 m primary mirror with 492 segments. With the high degree of segmentation, control of these segments against external disturbances to maintain the parabolic mirror shape becomes a challenging task. In this paper, a preliminary study of H ∞ control of the segmented primary mirror is presented. Both spatially-invariant distributed and centralized H ∞ controllers are designed for a proof-of-concept 37-segment system model. In our proposed mirror design, the segments are dynamically coupled through a uniform support structure, which is modeled using the finite element method. In addition, the segments have the coupled control objective of maintaining the overall mirror shape to allow the highest possible image resolution. A spatially-invariant distributed model is obtained by using a finite element model of the 7-segment system. The resulting model is then used to synthesize a spatially-invariant distributed controller, and is expanded to obtain the 37-segment system by interconnecting the distributed model in a honeycomb-like pattern. The simulation results of the distributed controller is presented and compared with the results from the centralized H ∞ controller applied to the 37-segment model. It is shown that both centralized and spatially-invariant distributed controllers satisfy the imaging performance requirements. © 2011 Elsevier Ltd. All rights reserved.

In order to achieve high optical collector and therefore high solar field efficiencies parabolic trough concentrators in concentrating solar power plants need to maintain their parabolic shape during operation. Additional to shape deviation already induced by the manufacturing process, deformation due to gravity load is an inevitable factor influencing shape accuracy in all types of parabolic trough collectors. This paper characterizes and quantifies the effect of gravity load on mirror shape and resulting slope and focus deviation values. One inner and one outer ideally parabolic shaped mirror of RP3 geometry mounted onto different support structures are evaluated in finite element analyses for all collector angles relevant for operation. The different support structures include two idealized support structures (ideal and elastic case) and one structure including relevant parts of EuroTrough type collectors (cantilever case). Constructional design and stiffness of the support structure significantly determine characteristic and magnitude of deformation. If compared to non-deformed shape, resulting rms values are as high as SDx = 1.7 mrad and FDx = 6.3 mm (inner mirror, elastic case) and SDx = 1.1 mrad and FDx = 5.6 mm (outer mirror, cantilever case). Depending on the type of support structure, minimum and maximum values occur at different collector angles. If compared to 0° (zenith) collector angle, resulting rms slope and focus deviation values are on average smaller than if compared to non-deformed mirror shape. This implies optimizing mirror shape for 0° (zenith) collector angle. However, it has to be considered that support structures for shape accuracy assessment in laboratory as well as support structures used in collector might differ significantly in design and stiffness, thus making it difficult to find one optimum shape for all types of mirror and collector.

We present an, omnidirectional vision system me have implemented to provide our mobile robot (Polyphemus"the one with only one eye) with a fast tracking capability. Polyphemus has to recognize elements in a semi-structured environment (the soccer field of Robocup) at a rate close to 20 Hz. We have designed a multi-shaped mirror to optimize the resolution of the image and its vertical coverage. The interpretation system is based on focused, multi-level, and opportunistic procedures.