中文摘要：

针对标准遗传算法收敛速度慢、易早熟，提出了一种基于进化阶段的动态加速自适应遗传算法，使遗传算法的搜索空间动态缩小并在遗传算法的选择算子中引入模拟退火机制，同时对交叉算子和变异算子做了自适应改进。将该算法应用于优选广东省韩江堂荆流域新安江模型的参数中，效果很好。

英文摘要：

Since standard genetic algorithm is slowly converged and has drawback of premature convergence, a dynamically accelerating-adaptive Genetic Algorithm based on evolution stages is presented, which can make the search space dynamic convengenee. In this algorithm, simulated anealing theory is adopted in the selection, and the operators of crossover and mutation is self-adptive. The results obtained from the application of this algorithm to Xinanjiang Model are satisfactorying.