提出一种基于小生境的GEP改进算法,将改进k-均值的聚类分析与遗传机制相结合,通过调节最小聚类距离,控制收敛的小生境数目,以提高算法跳出局部最优的能力.将改进算法应用在函数发现问题中并与基本GEP算法结果进行对比,实验表明改进算法具有更高的精度和更强的寻优能力.
This paper presents a hybrid GEP algorithm with niching, which combines a k-means clustering method and genetic mechanism; it tries to adjust the minimum clustering distance in order to decide the niching number and to avoid the problem of premature convergence. Experiments on three function funding problems show that the algorithm has higher precision and better search ability than basic GEP.