Optimal tilt and orientation maps: a multi-algorithm approach for heterogeneous multicore-GPU systems

摘要

This paper presents a new Geographic Information Systems (GIS) tool to compute the optimal solar-panel positioning maps on large high-resolution Digital Elevation Models (DEMs). In particular, this software finds out (1) the maximum solar energy input that can be captured on a surface located at a specific height on each point of the DEM, and then (2) the optimal tilt and orientation that allow capturing this amount of energy. The radiation and horizon algorithms we developed in previous works were used as baseline for this tool (Romero et al. in Comput. Phys. Com-mun. 178(11):800-808, 2008; Tabik et al. in Int. J. Geogr. Inf. Sci. 25(4):541-555, 2011). A multi-method approach is analyzed to make the hybrid implementation of this tool especially appropriate for heterogeneous multicore-GPU architectures. The experimental results show a high numerical accuracy with a linear scalability.