论文部分内容阅读
为了进一步提升随机漂移粒子群优化(RDPSO)算法的全局搜索能力、收敛速度以及在高维问题上的优化能力,提出一种基于频繁覆盖策略的RDPSO(FC-RDPSO)算法,并采用概率统计方法和蒙特卡罗方法分析频繁覆盖策略的可行性.在CEC’2013RPO的测试函数上将FC-RDPSO算法与多种优化算法进行对比,实验结果表明所提算法在收敛速度和全局搜索能力上表现出了突出的性能;在一组被广泛使用的大规模全局优化测试函数上的实验结果表明,FC-RDPSO算法在高维问题上同样表现出了较强的优化能力.
In order to further improve the global search ability, convergence speed and the optimization ability of the RDPSO algorithm, an RDPSO (FC-RDPSO) algorithm based on frequent coverage strategy is proposed, and the probability and statistics method And Monte Carlo method to analyze the frequent coverage strategy.Comparison of FC-RDPSO algorithm with many optimization algorithms in the test function of CEC’2013RPO, the experimental results show that the proposed algorithm in terms of convergence speed and global search ability The experimental results on a set of widely used large-scale global optimization test functions show that the FC-RDPSO algorithm also shows a strong optimization ability in high-dimensional problems.