参数线性化率定方法在HBV-IWS模型中的应用

摘要

Conventional objective function of minimum error square sum for nonlinear function parameter cal-ibration may increase unrelated local optimal value. A linearizied calibration method for non-linear function parameter has been developed in this paper, which is applied for parameter estimation of daily and hourly model for HBV-IWS. The results of the ideal daily model show that the method can converge the real val-ue. The performance of 13 years and 18 flood events observed data is satisfactory. The results show that this method is effective and reasonable and has quick convergence rate. Moreover,only sensitive parameters need to be calibrated, which can reduce the number of calibrated parameters and increase the efficiency of parameter calibration.%针对误差平方和最小为目标函数率定非线性函数参数会增加不相关的局部优值等问题，提出了参数线性化率定方法，并将其应用于HBV-IWS模型的日模拟和次洪参数优选中。首先通过日模拟理想模型的实例分析，发现参数均能收敛到真值，验证了该方法的可行性；然后对东张水库13年的实测日资料和18场洪水进行应用验证，结果表明，参数线性化率定方法在实际应用中是可行的、合理的，且寻找参数优值的速度较快；同时利用该方法率定敏感性参数，减少了率定所有参数带来的麻烦，提高了参数率定的效率，具有良好的应用前景。