Teaching-Learning-Based Differential Evolution Algorithm for Optimization Problems

Abstract

Differential Evolution (DE) is one of the current best evolutionary algorithms. It becomes the popular research topic in many fields such as evolutionary computing and intelligent optimization. At present, DE has successfully been applied to diverse domains of science and engineering, such as signal processing, neural network optimization, pattern recognition, machine intelligence, chemical engineering and medical science. However, almost all the evolutionary algorithms, including DE, still suffer from the problems of premature convergence, slow convergence rate and difficult parameter setting. To overcome these drawbacks, we propose a novel Teaching-Learning-Based Differential Evolution Algorithm(TLDE), in which the pheromone and the sensitivity model in free search algorithm to replace the traditional roulette wheel selection model, and introduces OBL to present an improved artificial bee colony algorithm. Experimental results confirm the superiority of Teaching-Learning-Based Differential Evolution Algorithm over several state-of-the-art evolutionary optimizers.