Development of methodology and software for detailed solar energy analysis in 3D

Creating 3D city models is a hot topic today. Achieving a well-defined 3D representation of urban environments requires large amounts of collected data combined with heavy processing of the data, both manually and automatically. The complexity makes it both expensive and time consuming to create accurate city models. An area which will truly benefit from well-defined city models is solar analysis. There are many tools available today for solar analysis but something all the tools have in common is that radiation calculations are very time consuming. The flexibility of the calculations in time and space domain are also below satisfaction. In this thesis a software was developed and included in a new methodology to go from high resolution elevation data to complete 3D environments. Solar radiation calculations were also implemented to be performed on the finished 3D buildings. The proposed method for radiation calculations is using the fact that the GPU handles some tasks better than the CPU and thereby reducing the calculation times. The results show that the finished 3D models are a relatively accurate representation of the real-life objects and they are well adapted for the purpose of radiation analysis. A validation of radiation calculations were also done using radiation sensors on modelled buildings. According to measured values the model seems to make a good estimation of clear-sky radiation. The calculation times are promising and there’s most probably room for further optimizations