Developing a dynamic model and software calculating solar energy potential using 3-dimensional data

Keywords: DSRM, GPU, solar energy, 3D buildings, sustainable enviroment Buildings play an important role in sustainable development. Performing solar radiation analysis on buildings at the multiple or individual scale is an important step for the formation of a sustainable environment. The accuracy of solar radiation analysis in buildings depends on the holistic an alysis of the buildings and the efficiency of the model. Traditional solar radiation analysis approaches use some technologies such as LIDAR, ALS, aerial photographs, satellite images, and MLS. These technologies are expensive and not suitable for non-existing structures. In this study, a dynamic solar radiation model (DSRM) has been developed to evaluate the global solar radiation on structures which can be either in planning stage or completed at the multiple or individual scale. Dynamical changes of shadow areas on building surfaces affect solar energy estimations. It is not enough to use only radiation estimation models for precise radiation analysis on buildings, different algorithms are required. In the proposed approach, finite element method, back-face detection and ray-tracing algorithms are utilized to obtain more precise results. Thus, real-time shadow analysis and analysis of the desired sensitivity and time interval can be obtained using the proposed approach. Finding the optimum value of passive parameters within the dynamics and complexity of an urban environment is a time consuming and difficult task. A new GPU-based radiation model has been developed to shorten the analysis time of DSRM implementation. GPU-DSRM is a flexible model that can be applied to any home plan or multiple home plans in a site. The GPU-DSRM software tool is implemented using the C CUDA architecture provided by NVIDIA. In this study, a site consisting of 10 buildings has been selected as a study area to verify the GPU-DSRM algorithm and to analyze its performance. Many tests with different scenarios on this area have been performed. In the experimental studies, the results obtained from both CPU and GPU based applications were compared and a 45-fold acceleration was achieved with GPU based application. With this new approach, solar potential analysis of building designs or existing urban formations can be completed in acceptable time.