白洋淀天然入淀水量在长期的时间序列上有着丰、枯水期交替演化的规律，灰色波形模型适用于这一规律发展趋势的研究。通过遗传算法（GA）对灰色一阶模型（GM（1,1））的迭代基值α与背景值系数β进行优化，利用遗传算法收敛效率高，选择范围广的优点，建立了以GA-GM(1,1)群为基础的GA-灰色波形模型，对白洋淀天然入淀水量趋势进行研究。最终得出结论：GA-灰色波形模型不仅在信息序列的拟合上明显优于传统灰色波形模型，且GA-灰色波形模型能更好的抓住信息序列发展特点，更为准确的预测白洋淀天然入淀水量演化规律。说明用GA-灰色波形模型进行白洋淀天然入淀水量研究是可行的，也为研究湖泊水资源量变化提供了一种新思路。

Seen from the long time-series of the nature flow into Baiyangdian Lake, it has a rule of the dry season and the wet season alternating. The grey wave forecasting model is applicable to research the rule’s evolutional trend. Using the genetic algorithm (GA) to optimize the first-order grey model（GM（1,1））’s iterative basic value α and background value coefficient β. Taking advantages of GA in efficient convergence and broad select range, basing on the GM(1,1) grey models established the GA-Grey wave forecasting model, thus use it study the trend of the nature flow into Baiyangdian Lake. The final analysis conclusion is that the GA-Grey wave forecasting model is obviously superior to the tradition grey wave forecasting model not only in matching information sequences but also in finding the changing characteristic of the information sequences, thus better forecast the changing of nature flow into Baiyangdian Lake. It shows the feasibility of GA-Grey wave forecasting model, providing the research of the variation of lake water resources quantity a new way.

国家自然科学基金创新研究群体科学基金资助项目(51321065)