Teaching-learning based optimization with crossover operation

Abstract

This paper developed a new variant of teaching-learning-based optimization (TLBO), called Teaching-Learning-Based Optimization with Crossover (TLBOC), for improving the performance of TLBO. The TLBOC incorporated the conventional crossover operation of differential evolution (DE) algorithm into teaching phases, which aims at balancing local and global searching effectively. Moreover, an estimation of distribution operation is used to predict a learning elitist. The learning elitist helps to boost learning efficiency of each student in learning phase. The performance of TLBOC is assessed for solving global unconstrained optimization functions with different characteristics. Compared to the TLBO and several other promising heuristic methods, numerical results reveal that the TLBOC has better optimization performance.