新安江模型参数全局优化研究

采用单纯多边形进化算法对3个气候与流域条件各不相同的流域的新安江模型计算参数优化问题进行了研究．结果表明：对于没有误差的水文资料，采用单纯多边形进化算法，可以优化出新安江模型的全部参数并使参数收敛到真值；对于实测的水文资料，由于模型参数之间的共线性，固定参数B与EX，可同时率定模型其他参数，随着实测资料系列的增加，率定参数值将趋于稳定．

Application of global optimization to calibration of Xin'anjiang model

A study was performed on the parameter optimization of the Xin'anjiang model for three watersheds with different meteorological and hydrogeographical conditions by use of the Schuffle Complex Evolution (SCE-UA) Algorithm. The result shows that, for the ideal hydrological data generated by the hydrological model, the real values of all the parameters of the Xin'anjiang model can be obtained by the SCE-UA optimization, and that for practical hydrological data, owing to the co-linearity between model parameters, the optimized parameters can be obtained by fixing of parameters B and EX, and with the increase of the hydrological data series, the optimized parameters converge to stable and real values.