论文部分内容阅读
为了实现全局最优的快速搜索,对全局最优所处的范围进行预估非常关键.受沙盘重心的力平衡关系启发,提出基于重心定位的全局最优邻域估计方法.利用空间变换技术,将寻优空间的重心“移”至全局最优附近;在重心逐次“移”动过程中,以重心为几何中心,构造一个范围不断缩小的全局最优邻域.函数测试和工程应用表明,所提出方法能以较小的种群和迭代次数确定全局最优的有效邻域范围.
In order to achieve the global optimal fast search, it is very important to estimate the scope of the global optimum.According to the force balance relationship of the gravity center of the sand table, a global optimal neighborhood estimation method based on the gravity center is proposed.Using spatial transformation technology, The center of gravity of the optimization space is “shifted” to the global optimum near the center of gravity; in the course of the gravity center “shift ” movement, the center of gravity is the geometric center, so as to construct a globally narrow optimal neighborhood with decreasing scope. Engineering application shows that the proposed method can determine the globally optimal effective neighborhood with smaller population and iteration times.