一个求解有约束的全局优化问题的变换函数法

Quasi-filled Transformation Function for Constrained Global Optimization

给出了求解一般的有约束非线性规划问题全局最优解的拟填充变换函数方法,而且讨论了所构造的变换函数的几个性质,按照其理论性质设计了一个变换函数算法,并进行了数值试验。数值实验表明,所给的方法是有效的。

A new quasi-filled transformation function with one parameter on constrained problem escaping the current local minimizer of global optimization is proposed. Firstly, the difinition of a quasi-filled transformation function for constrained problem is given and under certain assumptions, the proposed function is a quasi-filled transformation function is proved. Then a new algorithm according to the theoretical analysis is presented. Promising computational results are included and show the efficiency of the quasi-filled transformation function.