期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2018
卷号：96
期号：23
出版社：Journal of Theoretical and Applied
摘要：Conjugate Gradient Method (CG) is one of the well-developed gradient based method in solving optimization problems. It been widely used in solving large scale optimization problems due to its low computational cost and high efficiency in locating optimization solution. However, this method often fails to obtain global optimum solution when solving multimodal nonconvex optimization problems because once this method obtained a local optimum solution, it unable to move to another valley to obtain a better optimum solution. In this paper, ABCED Conjugate Gradient Method which consist of a series of enhanced conjugate gradient methods have been introduced to solve multimodal nonconvex optimization problems. The new developed methods have been tested with several benchmark problems. The numerical results had proved the effectiveness of the ABCED Conjugate Gradient Methods. The results showed the ABCED Conjugate Gradient with Fletcher-Reeves formula able to globally solve 80.95% of the selected benchmark test function. Then, ABCED Conjugate Gradient with Hestenes-Stiefel and Dai-Yuan formula had globally solved 76.19% of selected benchmark test function. However, ABCED Conjugate Gradient with Polak-Ribiere only able to solve one third of the selected benchmark test function.
关键词：Gradient Based Method; Conjugate gradient method; Artificial Bees Colony (ABC); Multimodal Non-convex Optimization; Global Optimization
