目前绝多数以误差平方和函数为参数估计的信息来源的参数估计方法存在诸多问题。介绍了一种在参数函数曲面上寻找最优参数的方法:函数曲面参数率定方法。并将该方法用于马斯京根模型的参数估计中。文中选用了几个例子进行参数率定,分析和比较的结果表明该方法比常规的方法精度高,而且该方法不受参数初值的影响,率定结果稳定,是一种有效的全局优值参数估计方法,也可用于其他非线性模型的参数估计。
Most of the current optimization algorithms try to find the optimum based on the objective function. A new method based on parameter function surface is introduced and used for optimizing the parameters of Muskingum routing model. The example analy- sis and comparison with other optimization algorithms show the method's good performance in iteration procedure, convergence and high precision accuracy. It's an effective global optimization algorithm and can be used for other nonlinear model.