一种带有附加记忆策略的改进教与学优化算法

An Improved TLBO Algorithm with Additional Memory Strategy

教与学优化算法是一种模拟班级教学现象的新型群体智能优化算法,算法参数简单,收敛速度快,已经在函数优化、工程计算等领域取得广泛应用。但是算法后期容易陷入局部收敛,为此提出了一种带有附加记忆策略的教与学优化(MTLBO)算法。该算法首先在教学阶段增加教师记忆策略,学生的历史记忆知识与教师历史教学能力对提高班级的整体教学水平具有重要的作用,在每次更新学习者的同时考虑教师上一代的最优值和当代的最优值,有效增强算法局部搜索能力;在学习阶段增加个体向最优个体和随机个体学习策略,多个学生互相学习,充分利用班级内的知识信息,从而增强了算法的全局搜索能力。采用具有不同特征的多个测试函数对算法进行仿真实验,并与基本TLBO算法和2种改进的TLBO算法进行对比分析,结果表明提出的MTLBO算法在获得较高的收敛精度和稳定性的同时还提高了收敛速度,有效避免算法局部收敛。

Teaching learning based optimization is a new type of swarm intelligence optimization algorithm that simulates class teaching phenomena.With simple parameters and fast convergence speed,the algorithm has been widely used in function optimization,engineering calculation and other fields.However,the algorithm tends to fall into local convergence later,so the modified teaching learning based optimization(MTLBO)with additional memory strategy is proposed.The teachers’memory strategy is added in the teaching stage and the students’historical memory knowledge and teachers’historical teaching ability play an important role in improving the overall teaching level of the class.When updating learners each time,the optimal value of the previous generation and the current optimal value of teachers are considered,effectively enhancing the local search ability of the algorithm.In the learning stage,the individual learning strategies are added to the optimal individual and random individual,so that multiple students can learn from each other and make full use of the knowledge information in the class,thus enhancing the global search ability of the algorithm.The proposed algorithm is simulated by multiple test functions with different characteristics,and compared with the basic TLBO algorithm and two improved TLBO algorithms.It is showed that the proposed MTLBO algorithm not only achieves higher convergence accuracy and stability,but also improves the convergence speed,effectively avoiding the local convergence of the algorithm.