传统优化算法和现代优化算法均是基于目标函数曲面寻求优值,而大多数参数率定方法的目标函数均为误差平方和。直观看来,基于误差平方和的目标函数曲面往往较参数函数曲面更为复杂。此外,这些算法常常无法证明寻得的优值是全局优值,而且计算过程较复杂,效率较低。鉴于此,提出函数曲面参数率定方法,并应用于SAC理想模型和实际模型的参数率定。首先,通过理想模型验证了该方法的可行性;其次,通过东张流域13年日资料和13场洪水验证了该方法的实际应用效果。结果表明,该方法在实际应用中是可行的、效率与精度皆较高,是一种有效的参数优选方法。
Both traditional optimization algorithms and modern optimization methods calibrate parameters based on the objective function surface,and most of the objective functions are error sum of squares. Intui-tively, objective function surface based on error sum of squares are more complex than parameter function surface. In addition,those algorithms are often unable to prove the optimal values what they found are glob-al optimal values, and the calculation procedure are complicated and inefficient. As a result, the parameter calibration method based on parameter function surface is put forward and applied to calibrate the parame-ters of SAC ideal model and actual model. First of all,the viability of the method was verified by the ide-al model. Then,the application effects of the method were verified in actual model through 13 years’daily hydrology datum and records of 13 floods in the Dongzhang River basin. The Results in this study indicate that the method is feasible, high efficiency and high precision in practical application. Therefore, it is a useful method of parameter optimization.