Improved Teaching-Learning Based Optimization for Global Optimization Problems

Teaching-learning-based optimization (TLBO) is a new population-based meta-heuristic algorithm. In this paper, a new variant of TLBO-Improved Teaching-Learning-based optimization (ITLBO) is developed for solving global optimization problems. The proposed ITLBO incorporates the position updating operation of swarm intelligent algorithm into different phases and aims at effectively balancing the local and global searching. Gaussian perturbation strategy is presented to prevent TLBO algorithm from trapping into local minima. Moreover, opposition-based learning technique is employed in learning phase to expand the exploration space. Experimental results reveal that ITLBO appear to enhance the solution accuracy and quality compared to TLBO and other promising heuristic methods.