期刊
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
卷 9, 期 2, 页码 199-215出版社
SPRINGER HEIDELBERG
DOI: 10.1007/s13042-017-0739-8
关键词
Bat algorithm; Random triangle-flipping strategy; Directing triangle-flipping strategy; Hybrid triangle-flipping strategy
资金
- National Natural Science Foundation of China [61663028]
- Natural Science Foundation of Shanxi Province [201601D011045]
- International Science AMP
- Technology Cooperation Program of China [2014DFR70280]
Bat algorithm (BA) is a novel population-based evolutionary algorithm inspired by echolocation behavior. Due to its simple concept, BA has been widely applied to various engineering applications. As an optimization approach, the global search characteristics determine the optimization performance and convergence speed. In BA, the global search capability is dominated by the velocity updating. How to update the velocity of bats may seriously affect the performance of BA. In this paper, we propose a triangle-flipping strategy to update the velocity of bats. Three different triangle-flipping strategies with five different designs are introduced. The optimization performance is verified by CEC2013 benchmarks in those designs against the standard BA. Simulation results show that the hybrid triangle-flipping strategy is superior to other algorithms.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据