二次分配问题是一个NP-hard问题,它在线路板设计、布局问题以及打字机键盘的设计等现实生活中有许多的应用。使用基本蚁群算法进行搜索时,其全局优化性能的优劣在很大程度上与蒸发系数的选择有关,若选择不合适,易使算法陷入局部最优。为此,本文提出一种新的算法,即将基本蚁群算法与禁忌搜索策略相结合来求解二次分配问题,设计出具体的算法模型,并对标准问题库中的具体实例进行测试,实验结果证实新方法的有效性。
Quadratic Assignment Problem(QAP) is a NP-hard problem and it has numerous real life applications in, for example, school layout, circuit board designing and typewriter keyboard designing. Ant Colony Algorithm(ACA) behaves well in finding local optium, whereas its global search depends on selection of the evaporation coefficient. An unsuitable evaporation coefficient may result in local optium of final solutions. So this paper presents a new Hybrid Ant Colony Algorithm (HACA) to solve quadratic assignment problem. The results obtained show that the new approach is efficient.