• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International MultiConference of Engineers and Computer Scientists >A Hybrid of Modified Simplex and Steepest Ascent Methods with Signal to Noise Ratio for Optimal Parameter Settings of ACO
【24h】

A Hybrid of Modified Simplex and Steepest Ascent Methods with Signal to Noise Ratio for Optimal Parameter Settings of ACO

机译：一种用于噪声比的改进的单纯形和最陡峭上升方法的混合，以获得ACO的最佳参数设置

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Metaheuristics are sequential processes that perform exploration and exploitation in the solution space aiming to efficiently find near optimal solutions with natural intelligence as a source of inspiration. One of the most well-known metaheuristics is called Ant Colony Optimisation, ACO. This paper is conducted to give an aid in complicatedness of using ACO in terms of its parameters: number of iterations, ants and moves. Proper levels of these parameters are analysed on eight noisy continuous non-linear continuous response surfaces. Considering the solution space in a specified region, some surfaces contain global optimum and multiple local optimums and some are with a curved ridge. ACO parameters are determined through Modified Simplex, MSM and Steepest Ascent methods, SAM, including their hybridisation. SAM was introduced to enhance a performance of MSM via the statistically significant regression analysis and Taguchi's signal to noise, S/N, ratio to recommend preferable levels of parameters. A series of computational experiments using each algorithm were conducted. Experimental results were analysed in terms of design points, best so far solutions, mean and standard deviation including S/N ratio. It was found that the results obtained from hybridisation were better than those using single algorithm itself. However, the average execution time of experimental run and number of design points using hybridisation were longer than those using a single method. Finally they stated a recommendation of proper level settings of ACO parameters for all eight functions that can be used as a guideline for future applications of ACO. This is to promote ease of use of ACO in real life problems.
机译：弥撒是顺序流程，在解决方案空间中进行探索和开发，旨在有效地发现附近的最佳解决方案与自然智能作为灵感的源泉。其中一个最着名的美术学习称为蚂蚁殖民地优化，ACO。本文进行了在其参数方面赋予使用ACO的复杂性：迭代，蚂蚁和移动的数量。在八个嘈杂的连续非线性连续响应表面上分析了适当的这些参数。考虑到指定区域中的解决方案空间，某些表面包含全局最佳和多个局部最优，有些表面具有弯曲脊。 ACO参数是通过修改的单纯形，MSM和最陡峭的Ascent方法，SAM确定，包括其杂交。引入SAM通过统计上显着的回归分析和Taguchi信号对噪声，S / N，比率来增强MSM的性能，以建议优选的参数。进行了一系列使用每种算法的计算实验。在设计点，最佳解决方案，均值和标准偏差方面分析了实验结果，包括S / N比。发现从杂交获得的结果优于使用单算法本身的结果。然而，使用杂交的实验运行的平均执行时间和使用杂交的设计点的数量比使用单一方法的实验点的数量长。最后，他们向所有八个函数表示了ACO参数的适当级别设置的推荐，这些功能可用作ACO的未来应用的指导。这是为了促进现实生活中的ACO的易用性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号