1、2023-05-10计算机应用,Journal of Computer Applications2023,43(5):1355-1364ISSN 1001-9081CODEN JYIIDUhttp:/基于协同变异与莱维飞行策略的教与学优化算法及其应用高昊,张庆科*,卜降龙,李俊青,张化祥(山东师范大学 信息科学与工程学院,济南 250358)(通信作者电子邮箱)摘要:针对教与学优化(TLBO)算法在处理优化问题时存在搜索不均衡、易陷入局部最优、综合求解性能弱等缺陷,提出一种基于均衡优化与莱维飞行策略的改进教与学优化算法ELMTLBO。首先设计精英均衡引导策略,通过种群中多个精英个体的均衡引导提
2、高算法的全局寻优能力;其次在TLBO算法的学习者阶段后,利用自适应权重策略对莱维飞行产生的步长进行自适应缩量,以提高种群局部寻优能力,增强个体对复杂环境的自适应性;最后设计了变异算子池逃逸策略,通过多个变异算子的协同引导,提升算法的种群多样性。为验证算法改进的有效性,将EMLTLBO算法与侏儒猫鼬优化算法(DMOA)等先进的智能优化算法以及平衡教与学优化(BTLBO)算法、标准TLBO等同类型算法在15个国际测试函数上进行综合收敛性能比较。统计实验结果表明,与先进的智能优化算法和TLBO算法变体相比,ELMTLBO算法能够有效平衡其搜索能力,不但有效求解单峰和多峰问题,而且在复杂多峰问题上仍有
3、显著的寻优能力。在不同策略的共同作用下,ELMTLBO 算法的综合优化性能突出,全局收敛性能较为稳定。此外,ELMTLBO算法成功应用于基于隐马尔可夫模型(HMM)的多序列比对(MSA)问题中,优化后得到的高质量对齐序列可用于疾病诊断、基因溯源等,可为生物信息学提供算法支撑。关键词:教与学优化算法;均衡引导;莱维飞行;自适应权重;变异算子池;隐马尔可夫模型;多序列比对中图分类号:TP18;TP391 文献标志码:ATeaching-learning-based optimization algorithm based on cooperative mutation and Lvy flight
4、 strategy and its applicationGAO Hao,ZHANG Qingke*,BU Xianglong,LI Junqing,ZHANG Huaxiang(School of Information Science and Engineering,Shandong Normal University,Jinan Shandong 250358,China)Abstract:Concerning the shortcomings of unbalanced search,easy to fall into local optimum and weak comprehens
5、ive solution performance of Teaching-Learning-Based Optimization(TLBO)algorithm in dealing with optimization problems,an improved TLBO based on equilibrium optimization and Lvy flight strategy,namely ELMTLBO(Equilibrium-Lvy-Mutation TLBO),was proposed.Firstly,an elite equilibrium guidance strategy w
6、as designed to improve the global optimization ability of the algorithm through the equilibrium guidance of multiple elite individuals in the population.Secondly,a strategy combining Lvy flight with adaptive weight was added after the learner phase of TLBO algorithm,and adaptive scaling was performe
7、d by the weight to the step size generated by Lvy flight,which improved the populations local optimization ability and enhanced the self-adaptability of individuals to complex environments.Finally,a mutation operator pool escape strategy was designed to improve the population diversity of the algori
8、thm by the cooperative guidance of multiple mutation operators.To verify the effectiveness of the algorithm improvement,the comprehensive convergence performance of the ELMTLBO algorithm was compared with 7 state-of-the-art intelligent optimization algorithms such as the Dwarf Mongoose Optimization
9、Algorithm(DMOA),as well as the same type of algorithms such as Balanced TLBO(BTLBO)and standard TLBO on 15 international test functions.The statistical experiment results show that compared with advanced intelligent optimization algorithms and TLBO algorithm variants,ELMTLBO algorithm can effectivel
10、y balance its search ability,not only solving both unimodal and multimodal problems,but also having significant optimization ability in complex multimodal problems.It can be seen that with the combined effect of different strategies,ELMTLBO algorithm has outstanding comprehensive optimization perfor
11、mance and stable global convergence performance.In addition,ELMTLBO algorithm was successfully applied to the Multiple Sequence Alignment(MSA)problem based on Hidden Markov Model(HMM),and the high-quality aligned sequences obtained by this algorithm can be used in disease diagnosis,gene tracing and
12、some other fields,which can provide good algorithmic support for the development of bioinformatics.文章编号:1001-9081(2023)05-1355-10DOI:10.11772/j.issn.1001-9081.2022030420收稿日期:2022-04-01;修回日期:2022-05-20;录用日期:2022-05-30。基金项目:国家自然科学基金资助项目(62006144,62176144,61772322);山东省自然科学基金重大基础研究项目(ZR2019ZD03);山东泰山学者计
13、划项目(ts20190924)。作者简介:高昊(1996),男,山东淄博人,硕士研究生,CCF会员,主要研究方向:进化计算、群体智能;张庆科(1985),男,山东济宁人,博士,CCF会员,主要研究方向:群体智能、进化计算;卜降龙(1998),男,山东泰安人,硕士研究生,CCF会员,主要研究方向:进化计算、群体智能;李俊青(1976),男,山东聊城人,副教授,博士,主要研究方向:智能优化和调度;张化祥(1966),男,山东济宁人,教授,博士,主要研究方向:机器学习、模式识别、进化计算。第 43 卷计算机应用Key words:Teaching-Learning-Based Optimizatio
14、n(TLBO)algorithm;equilibrium guidance;Lvy flight;adaptive weight;mutation operator pool;Hidden Markov Model(HMM);Multiple Sequence Alignment(MSA)0 引言 近年来,随着待优化实际问题日益复杂,各种智能优化算法为代替传统优化方法而相继被提出,具有效率高、求解代价小的特点,以智能随机的方式求解,在计算精度、收敛速度、单目标与多目标问题、单峰与多峰等问题上各有侧重。教与学优化(Teaching-Learning-Based Optimization,TLBO
15、)算法1是一种针对连续非线性大规模优化问题的优化算法,无需提前设定参数,不必关注参数对解质量的影响。目前,许多文献已将 TLBO 算法及其相关变体应用于比例-积分-微分控制器(Proportion Integration Differentiation,PID)优化2、系统优化3、风险预测4、路径优化5、序列比对6、特征选择7、经济负荷调度8、故障检测9、图像工程10等领域,并取得了不错的应用效果。与众多智能优化算法一样,TLBO算法也会有陷入局部极值、算法早熟等问题。为了更好地解决这些问题,本文对TLBO算法相关研究进展归纳总结如下:1)种群初始化。Wang 等11使用 Logistic 混
16、沌映射初始化种群,生成均匀分布在解空间的个体,提升了种群多样性;Roy等12提出一种基于反向学习的教与学优化(Oppositional TLBO,OTLBO)算法,使用反向学习生成反向种群,并在算法迭代过程中产生反向学习种群,与种群中相对应的个体进行比较,筛选比较好的个体来加快算法收敛速度。2)引入策略和机制。Chen等13提出了一种可变种群方案,种群规模随三角形震荡折线图的波动动态变化,在种群规模增长阶段使用高斯分布生成新个体,在种群规模减小阶段删除最相似的个体;Yu等14提出了一种分组策略,将种群分为若干组进行老师的教学任务,防止算法过早收敛,同时在学生交流阶段让某位学生随机选择两名学生进行交流,以提升算法的种群多样性;Wang等15让学习者在迭代学习过程中使用差异变异来保持学习者的多样性,避免了采用重复消除近似个体的重启方法造成算法时间复杂度的提升;Taheri 等16提出了一种平衡教与学优化(Balanced TLBO,BTLBO)算法,新增了辅导阶段与重启阶段,在提升算法局部勘探能力的同时又能兼顾全局探索能力;He等17提出了一种混沌教与学优化(Chaotic TLBO,CT