求解约束优化问题的改进蝙蝠算法

Modified bat algorithm for solving constrained optimization problems

针对基本蝙蝠算法求解精度低、易陷入局部最优的缺点,提出一种改进的蝙蝠算法用于求解约束优化问题。该算法利用佳点集方法构造初始种群以维持群体的多样性,引入惯性权重以协调算法的勘探和开发能力。为了避免算法陷入局部最优,对当前全局最优解进行多样性变异操作。通过对四个标准测试函数和化工应用的仿真实验并与其他算法进行比较,结果表明了该算法具有较强的全局搜索能力。

Bat algorithm（ BA） has a few disadvantages in the global searching,including low solving precision and high possibility of being trapped in local optimum. This paper proposed a modified bat algorithm for solving constrained optimization problems. It used good-point set method to initiate individual position and velocity,which strengthened the diversity of global searching. It introduced the inertia weight to coordinate the algorithm＇s ability of exploration and exploitation. The proposed modified algorithm was tested on four standard benchmark functions and one chemical engineering optimization problem. Experimental results indicate that the proposed algorithm is able to find better solutions comparable to the chosen methods.