基于遗传算法的水库中长期调度函数研究

Research on the Function of Reservoir Long-term Operation Based on Genetic Algorithms

传统采用对确定性优化算法所得的运算结果进行回归分析而建立水库调度函数的方法,对原有结果的拟合有较大的偏差;为此,建立水库线性调度函数求解模型,并用改进的遗传算法进行求解,同时采用“培训”的方法,加快算法收敛速度及性能。与基于POA优化的回归结果对比分析表明,遗传算法求解调度函数具有优势。

In traditional way, we can obtain the function of reservoir operation by using regression analysis, from the course of reservoir operation which can be obtained by some optimization methods; but the result becomes bad, sometimes not accepted; this paper establishes the model of solution to linear function of reservoir operation, and then introduces a solution to this model based on advanced genetic algorithms. The training brought in the method can improve genetic algorithms effectively. Comparing with the regression result based on POA method, superiority is apparent.