基于蚁群全局优化的二层规划模型与仿真

Two-layer Plan Model& Simulation Based on Ant Colony Global Optimization

二层规划问题通常是一个非凸问题,因此在实际工程领域对其进行求解具有极大的困难。研究了问题的本质特征,提出了一种组合设计算法模型——基于蚁群算法求解二层规划问题的全局优化策略。组合算法采用蚁群算法求解上层问题,下层的线性规划问题则采用单纯型算法完成求解。设计的组合算法思路清晰,仿真计算结果表明,该算法有着良好的全局收敛可靠性和较高的收敛速度,是目前求解此类两层线性规划问题的一种有效算法。

Two-layer plan problem is usually a non-convex problem, so it is very difficult in solving practical engineering problem. By fully study the essential characteristics of the problem, this paper presents an algorithm based on the ant colony for solving a class of two-layer linear programming global optimization problems. We can use Ant colony algorithm for solving the upper problem and simplex algorithm for the lower the linear progranmming problem. The combi- nation of algorithms design is very clear. The numerical results show that the algorithm has good global convergence reliability and high convergence speed, so it is efficient for solving a two-layer linear prograraming problems.