【摘要】 本文提出了一种快速求解全局优化问题的跨越函数方法,与以填充函数法为代表的一类全局优化方法相比,本文定义的跨越函数直接凸显了在求解全局优化问题时构造辅助函数的目的,更重要的是跨越函数方法能够一步跨过函数值比当前局部极小值高的区域,而直接找到原函数f(x)的位于函数值比当前局部极小值低的区域中的局部极小点,加快了全局寻优的过程,并且通过有限次迭代,找到全局最优解.更多还原

【Abstract】 In this paper,we propose a cross function method (CFM) for global optimization.Compared with the concept of the auxiliary function methods represented by the filled function method (FFM),the concept of the CFM shows the objective to construct the auxiliary functions more directly.Most importantly,CFM can help us to cross the higher region by only one step and immediately achieve the local minimizer of the previous function f(x) located in the lower region.Thus the process of global optimization is accelerated.The two regions is defined compared with the current local minimum.After finite iterations,the global minimizer is found.更多还原