论文部分内容阅读
为适应不断提高的震害预测需求,精细化模型已经成为城市区域震害预测的主要发展方向。然而传统的基于CPU平台的计算方法成本过高,使得精细化模型的应用受限。近年来,GPU技术由于其强大的并行计算能力和较为低廉的价格优势,在通用计算领域得到了快速的发展和应用。基于GPU-CPU协同计算技术,建立了城市区域震害的计算模拟方法,显著缩短了城市区域震害预测的时间。并采用该方法对某中型城市的真实震害进行了模拟应用,展示了GPU技术在大区域城市建筑群震害模拟中的独特优势。
In order to meet the increasing demand for earthquake damage prediction, the refined model has become the main development direction of earthquake damage prediction in urban areas. However, the traditional calculation method based on CPU platform is too expensive, which makes the application of the refined model limited. In recent years, GPU technology has been rapidly developed and applied in the field of general computing due to its powerful parallel computing power and relatively low price advantage. Based on the GPU-CPU collaborative computing technology, a computational simulation method of earthquake damage in urban areas is established, which significantly shortens the prediction time of earthquake damage in urban areas. This method is applied to simulate the real damage in a medium-sized city, and shows the unique advantages of GPU technology in the simulation of earthquake damage in large urban buildings.