一种优化精度较高的小生境遗传算法

提出了一种基于排挤机制的小生境遗传算法．其特点是能够在进化过程中动态地记录个体的进化方向，由此在进化结束后，可以沿记录的方向继续进行局部搜索，以进一步提高解的精度，仿真实验表明，该算法不但可以有效地保持种群的多样性．找到较多的解，还可大大提高解的精度，与普通遗传算法相比，虽然它有较大的时间代价，但在一些对快速性要求不高、而对解的精度要求较高的场合，有很好的应用前景。

A Novel Niche Genetic Algorithm of Accurate Optimization Performance

Based on crowding mechanism, a novel niche genetic algorithm is proposed which can record evolutionary direction dynamically during evolution. After evolution completed, the local searching along the recorded direction can greatly improve solutions precision. Simulation shows that this algorithm can not only keep population diversity but also find accurate solutions. Although the algorithm may take much time than standard GA, it is really worth applying to cases that are not demanding for time-consuming while demanding for high solution precision.