An advanced teaching-learning-based algorithm to solve unconstrained optimization problems

Abstract

The Teaching-Learning-Based Optimization (TLBO) algorithm is being extended to a broader range of applied optimization problems in the literature, mimicking the teaching-learning process. This paper proposes an Advanced Teaching-Learning-Based Optimization (Ad-TLBO) algorithm to enhance the efficiency and performance of the original version of TLBO in terms of accuracy, convergence rate, and reliability characteristics. The advancement is obtained by modifying the initialization, search approach, and structure of the two main phases of this algorithm in four steps to improve exploration and exploitation capability. Efficiency comparisons are shown in four challenges with various benchmark functions with multimodal, separable, differentiable, and continuity characteristics. The results are compared with several intelligent optimization algorithms. It is also deduced that this algorithm outperforms all investigated optimization algorithms in terms of accuracy, convergence speed, and success to reach acceptable solutions for various benchmark functions.