基于GA-灰色波形预测模型的白洋淀天然入淀水量

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。

Analyses on the nature flow into Baiyangdian Lake based on the GA-grey wave forecasting model

In terms of long time-series,the natural flow into Baiyangdian Lake has a pattern of alternating between dry season and wet season.The grey wave forecasting model is suitable for studying the development trend of this pattern.We used 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 advantage of GA's characteristics of efficient convergence and broad selection range,we established the GA-grey wave forecasting model based on the GA-GM(1,1) grey models,and used it to study the trend of the natural flow into Baiyangdian Lake.It was concluded that the GA-grey wave forecasting model is obviously superior to the traditional grey wave forecasting model not only in matching information sequences but also in finding the changing characteristics of the information sequences;thus it can better forecast the evolution of the natural flow into Baiyangdian Lake.This study has proved the feasibility of GA-grey wave forecasting model,and can provide a new thought to the research of the variation of water resources quantity of lakes.