| 摘要: |
| 泥石流具有突发性、速度快、破坏力强的特点。研究其运动规律和危险性预测,具有重要的现实意义和理论价值。本文发展了一套基于GPU(图形处理单元)加速技术的二维水动力模型,方程采用二阶Godunov格式有限体积方法求解,以确保计算精度,并使用GPU加速技术来提高计算效率。以溃坝型水槽泥石流运动特征及贵州省普洒村滑坡泥石流危险范围预测作为模拟参照,利用所构建的GAST(GPU Accelerated Surface Water Flow and Transport Model)模型完成泥石流水槽试验模拟和危险范围预测,结果表明数值模拟与实测结果吻合度较高；GPU加速技术的应用有效地解决了泥石流模拟过程中计算量大的问题。在PC机上模型运行快速高效。由此来看,此模型为泥石流运动规律模拟与预测提供了新方法,为放在减灾提供了技术支撑。 |
| 关键词: 泥石流 GPU加速计算 数值模拟 |
| DOI： |
| 分类号:P642. 23 |
| 基金项目:国家自然科学基金(52079106、51609199)；国家重点研发计划项目(2016YFC0402704)；作者简介：康永德(1989-),男,博士生,主要从事计算水力学数值模拟研究,E-mail: 2631310516@qq.com通信作者：侯精明(1982-),E-mail：jingming.hou@xat.edu.cn |
|
| Simulation of debris flow and hazard assessment based on GAST method |
|
Kang Yongde①1, Hou Jingming①1, Hu Chunhong②2, Baoshan Shi①1, Zenglin Hong③3
|
|
1.xi an university of technology;2.China Institute of Water Resources and Hydropower Research;3.Shaanxi GeologicalSurvey
|
| Abstract: |
| The debris flow has the characteristics of suddenness, fast flow speed and strong destructive force. It is of great practical significance and theoretical value to study its dynamic mechanism and accurately simulate the formation process. Model is within the second order Godunov-type finite volume scheme to preserve the accuracy of the simulation and employs graphic processing units (GPU) computing technique to accelerate the calculation. Taking the dam-breaking flume debris flow movement characteristics and the prediction of the risk range of the landslide and debris flow in Pusa Village, Guizhou Province as a simulation reference, the constructed GAST model was used to complete the simulation of the debris flow tank test and the prediction of the risk range, the results show that the numerical simulation is in good agreement with the measured results; the application of GPU acceleration technology effectively solves the large amount of calculation in the process of debris flow simulation Problem, the model runs fast and efficiently on the PC. It shows that the model is an ideal tool to study the sudden debris flow process, it can provide timely and accurate decision support for disaster prediction. |
| Key words: Debris flow GPU accelerated calculation Numerical Simulation |
