以长江三峡水库寸滩、武隆入库径流预报项目为依托,研究洪水预报模型参数率定。针对以往洪水预报模型参数率定所采用的人工试错法率定参数困难、精度不高的问题,提出粒子群优化算法和交互式率定方法。实例计算表明,算法加快参数优选速度,交互式率定提高了模型预报精度。
This paper is an attempt to do research on the parameter calibration of hydrology model by using the floodwater prediction at Cuntan and Wulong in Three Gorges Reservoir as the background. By reason of inefficient parameters calibration, a method based on particle swarm and interactive calculation is proposed. The application of this proposed model shows that it can accomplish parameter calibration rapidly and improve prediction accuracy, parameter calibration rapidly and find out the optimal solutions of parameters accurately.