A teaching-learning-based optimization algorithm with producer-scrounger model for global optimization

In order to decrease the computation cost and improve the global performance of the original teaching-learning-based optimization (TLBO) algorithm, the area-copying operator of the producer-scrounger (PS) model is introduced into TLBO for global optimization problems. In the proposed method, the swarm is divided into three parts: the producer, scroungers and remainders. The producer is the best individual selected from current population and it exploits the new solution with a random angle and a maximal radius. Some individuals, which are different from the producer, are randomly selected according to a predefined probability as scroungers. The scroungers update their position with an area-copying operator, which is used in the PS model. The remainders are updated by means of teaching and learning operators as they are used in the TLBO algorithm. In each iteration, the computation cost of the proposed algorithm is less than that of the original TLBO algorithm, because the individuals of the PS model are only evaluated once and the individuals of the TLBO algorithm are evaluated two times in each iteration. The proposed algorithm is tested on different kinds of benchmark problems, and the results indicate that the proposed algorithm has competitive performance to some other algorithms in terms of accuracy, convergence speed and success rate.