求解无约束优化问题的改进教与学优化算法

摘要

TLBO is a novel swarm intelligence optimization algorithm. Since the shortcoming of TLBO, a new improved teaching-learning-based optimization algorithm (ITLBO) is proved to solve unconstrained optimization problems. Firstly,a new method is a-dopted to define the average level of the students in the"teacher stage". Then,a linear decreasing inertia weigh factor is added in the"teacher" and "student" stage. Finally,the crossover operation is performed dynamically according to the fitness value with an adap-tive crossover operator in the algorithm. Through 11 unconstrained optimization problems are compared and tested,the results show that ITLBO is significantly better than the other four types of TLBO algorithm.%教与学优化算法(TLBO)是一种新型的群智能优化算法.针对算法求解性能的不足,对其进行改进并用于求解无约束全局优化问题.首先,在算法的"教师阶段"采用一种新的策略对学生平均水平进行定义,然后,在算法的"教师阶段"和"学生阶段"分别加入一种线性递减的惯性权重因子,最后,在算法中加入一种自适应精英交叉算子,不同粒子根据适应度值而动态执行交叉操作.通过11个无约束优化问题进行对比测试实验,结果显示,改进后的算法(ITLBO)在探索性能和收敛速度方面优于TLBO等其它四种类型的算法.