非线性混合整数规划的罚函数解法

A Method of Penalty Function for Solving Nonlinear Mixed Integer Programming

文[6]中,我们对非线性混合整数规划的解法进行了探讨,利用罚函数把有约束非线性混合整数规划问题化为等价的无约束非线性混合整数规划问题,然后把离散整变量连续化,从而非线性混合整数规划化为与之等价的无约束非线性规划。本文弱化了文[6]中定理1的条件,并得到了相应的结论。

In [6], a penalty function are proposed for nonlinear mix-integer programming （NMIP）, （NMIP） problems with bounded variables are transformed into equivalent nonlinear programming （NP）problems, and then the optimal solution about （NMIP） can be obtained by solving an unconstrained non-linear programming problem. In this paper, the condition of the Theorem 1 in [6] is reduced and the corresponding conclusion is obtained.