(488g) A Model-Based Feedback Optimization Method for Solar Dish Facets Alignment
This work attempts to provide a systematic model-based feedback optimization method to align mirror facets instead of trial-and-error or empirical methods which has long information cycle and inefficient use of data obtained through alignment. The proposed method uses as projected light flux density data in front of the focus point to guide the alignment of the mirror facet in real-time without any special light patterns or complex equipment. The method first build a rigid body model of the entire solar dish assembly as if it is perfectly aligned, then based on the model predicted flux density distribution off focus and observed flux density distribution to identify a series of off-focus hot spots sorted by size. Then for each hot spot, flux density distributions at different distances are used to locate the misaligned mirror facet(s) primarily responsible for the hot spot and compute adjustments needed, adjustments on the targeted facet(s) is then adjust accordingly to remove the hot spot. The procedures are then repeated until all hot spots are resolved, which ensure near perfect alignment of the entire solar dish assembly. A three layer solar dish consists of inner, middle, and outer ring is investigated as an example, simulation showed successful alignment of solar dish facets to parabolic shape following this method and experimental study also confirm the effectiveness of the proposed strategy under realistic gravity induced shape deform.