求解混合整数规划问题的指数变差积分算法

An exponent deviation integral algorithm for mixed integer programming

研究了一种求解混合整数规划问题的指数变差积分算法.利用积分型总极小值理论及指数变差积分对混合整数规划问题进行研究,通过变差积分函数的分析性质及混合整数规划的最优性条件,结合牛顿法设计了一种求解混合整数规划的指数变差积分新算法.运用Monte-Carlo模拟方法实现整个算法,数值结果表明该算法是有效的.

This paper studies an exponent deviation integral approach to the mixed integer programming problem. A deviation integral function with good properties is proposed, and the optimality condition for a mixed integer programming problem is examined. Then an exponent deviation integral algorithm is developed. Numerical calculation is performed using the Monte-Carlo technique to show effectiveness and feasibility of the algorithm.