Modified teaching-learning-based optimization algorithm

In allusion to the shortcoming of global exploration performance of basic Teaching-Learning-Based Optimization (TLBO) algorithm in solving complex high-dimensional problems, a modified Teaching-Learning-Based Optimization (MTLBO) algorithm is proposed. In MTLBO algorithm, the methods of teaching phase and learning phase are respectively modified to enhance to disturbance potential of search space, and a new “Self-Learning” method is presented to enhance the innovation ability of the learner and the global exploration performance. Finally, the performance of the proposed MTLBO algorithm is investigated over 6 complex high-dimensional benchmark functions. The results show that the proposed MTLBO algorithm has some advantages over convergence velocity, accuracy, and stability.