¹School of Computer Science and Technology, Xuchang University, Xuchang, Henan, 461000, P.R. China
²School of Computer Science, China University of Geosciences, Wuhan 430074, P.R. China

---

Abstract—Differential evolution (DE) is a simple yet powerful evolutionary algorithm for global numerical optimization. In this paper, we propose a novel hybrid DE variant to accelerate the convergence rate of the classical DE algorithm. The proposed algorithm is hybridized with a convex mutation. The convex mutation is able to utilize the information of the parents, and hence, provides faster convergence speed. Our proposal is referred to as Convex- DE. In order to verify our expectation, we test our approach on 13 widely used benchmark functions. The results indicate that our approach is better than the classical DE algorithm in terms of the convergence speed and the quality of final solution. Furthermore, the potential of our approach for real-world applications is evaluated on three real-world problems.

Index Terms—Differential evolution; convex mutation; numerical optimization; real-world applications.

[PDF]

Cite: Jingfeng Yan, Chaofeng Guo, Wenyin Gong, "Hybrid Differential Evolution with Convex Mutation," Journal of Software vol. 6, no. 11, pp. 2321-2328, 2011.