基于遗传程序设计的水库年末水位预测模型

为了克服传统预测方法的不足，采用遗传程序设计，建立了多年调节水库年末消落水位预测模型。该模型利用演化计算自动寻找最优的模型结构，比传统方法具有较大的灵活性和智能性。通过对洪家渡多年调节水库的实例计算，不仅得出影响因子与年末消落水位的最佳函数式，而且用于拟合和预报的精度均满足要求，可以为水库调度提供一定的依据。

Genetic Programming-based Prediction Model for the Year-end Reservoir Level

In order to overcome the disadvantages of traditional prediction methods, genetic programming is used to predict the year-end level of overyear regulation reservoirs. The model, which is more flexible and intelligent than traditional methods, can find out the optimal model structure automatically with evolvement calculation. With the Hongjiadu overyear regulation reservoir as an example, not only an optimal function of the influencing gene and the year-end level is obtained, but also the precision for fitting and forecast is satisfactory, providing a reference for reservoir scheduling.