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, 隐马尔科夫模型, 多序列比对

Abstract: Abstract: To address the shortcomings of the Teaching-Learning-Based optimization (TLBO) in dealing with optimization problems such as unbalanced search, easy to fall into local optimum, and weak comprehensive solution performance, a new teaching optimization algorithm (ELMTLBO) based on balanced optimization and the Lévy flight strategy is proposed in this paper. Firstly, the algorithm designs an elite balanced guidance strategy to improve the global merit-seeking ability of the algorithm through the equilibrium bootstrap of multiple elite individuals in the population. Secondly, a strategy combining Lévy flight with adaptive weights is added after the learner phase of the TLBO algorithm, and the weights perform adaptive scaling of the step size generated by the Lévy flight, which improves the population's local merit-seeking ability and enhances the self-adaptability of individuals to complex environments. Finally, a mutation operator pool escape strategy is designed, and the population diversity of the algorithm is significantly improved by the cooperative guidance of multiple mutation operators. To verify the effectiveness of the algorithm improvement, the paper compares the comprehensive convergence performance of the EMLTLBO algorithm with 7 state-of-the-art algorithms such as the Dwarf Mongoose Optimization Algorithm (DMOA) and 7 algorithms of the same type, such as the Balanced Teaching-Learning-Based Optimization Algorithm (BTLBO) and the Standard TLBO, on 15 international test functions. Statistical experimental results show that compared with algorithms such as TLBO, the convergence speed of the ELMTLBO algorithm on simple unimodal and multimodal problems is improved by more than 10% on average, and the success rate of finding the global optimal solution is improved by more than 50% on average. On complex multimodal problems, the convergence accuracy of ELMTLBO is improved by more than 1E-4 orders of magnitude on average compared to the better performing CTLBO. With the combined effect of different strategies, the algorithm has outstanding comprehensive optimization performance and more stable global convergence performance. In addition, this paper also successfully applies the ELMTLBO algorithm to the Multiple Sequence Alignment (MSA) problem based on the Hidden Markov Model (HMM), and the high-quality aligned sequences obtained by this algorithm can be used in disease diagnosis, gene tracing, virus control, and other fields, which can provide good algorithmic support for the development of bioinformatics.

Key words: teaching-learning-based optimization(TLBO), balanced guidance, levy flight, adaptive weights, mutation operator pool, hidden markov model(HMM), multiple sequence alignment(MSA)