Opposition-based learning grey wolf optimizer for global optimization

摘要

Grey wolf optimizer is a novel swarm intelligent algorithm. It has received lots of interest from the heuristic algorithm community for its superior optimization capacity and few parameters. However, it is also easy to trap into the local optimum when solving complex and multimodal functions. In order to boost the performance of GWO, an opposition-based learning grey wolf optimizer (OGWO) is proposed. The opposition-based learning approach is incorporated into GWO with a jumping rate, which can help the algorithm jump out of the local optimum and not increase the computational complexity. What is more, the coefficient a→ is dynamically adjusted by the nonlinear function to balance exploration and exploitation. The serial experiments have revealed that the proposed algorithm is superior to the conventional heuristic algorithms, it is also better than GWO and its variants.