引力搜索算法的改进研究

发布时间：2018-12-21 09:14

【摘要】：随着人类文明的不断进步与发展,优化问题已是人类生活中不可或缺的一部分。为了更好的解决此类问题,越来越多的学者致力于优化算法的研究。近年来,多种新型的启发式优化算法已被提出并能更好地解决复杂的最优化问题。其中,引力搜索算法(Gravitational Search Algorithm,GSA)在2009年由Rashedi首次提出,是一种新的启发式优化算法。该算法的改进与实际应用方面是学者们研究的主要工作。为了使引力搜索算法的优化能力达到最佳,对该算法的改进工作需要不断探索与研究。因此,本论文介绍了两种对于引力搜索算法的改进方法,进一步加强了该算法的优化性能。主要工作有:1、针对引力搜索算法易陷入局部最优的缺点,提出一种惯性质量衰减的引力搜索算法。惯性质量的衰减率由隶属度函数定义,并给出一个新的变异算子。最后,把所提出算法应用到经典测试函数中,并与标准引力搜索算法及其他改进的引力搜索算法比较,数值结果表明所给出的算法能够提高求解精度和收敛速度。2、针对引力搜索算法的早熟收敛现象,提出一种自适应变异的引力搜索算法。该算法将模拟退火算法(SA)的Metropolis准则和自适应变异概率的混合变异策略与引力搜索算法结合,数值实验表明,改进后的算法具有更好的性能。
[Abstract]:With the continuous progress and development of human civilization, optimization has become an indispensable part of human life. In order to solve this kind of problem better, more and more scholars devote themselves to the research of optimization algorithm. In recent years, a number of new heuristic optimization algorithms have been proposed and can better solve complex optimization problems. The Gravity search algorithm (Gravitational Search Algorithm,GSA), first proposed by Rashedi in 2009, is a new heuristic optimization algorithm. The improvement and practical application of the algorithm is the main work of scholars. In order to optimize the optimization ability of gravitational search algorithm, the improvement of the algorithm needs to be explored and studied continuously. Therefore, this paper introduces two improved methods for gravitational search algorithm, which further enhance the optimization performance of the algorithm. The main works are as follows: 1. Aiming at the disadvantage that gravity search algorithm is easy to fall into local optimum, a gravity search algorithm for inertial mass attenuation is proposed. The attenuation rate of inertial mass is defined by membership function and a new mutation operator is given. Finally, the proposed algorithm is applied to the classical test function and compared with the standard gravity search algorithm and other improved gravitational search algorithms. The numerical results show that the proposed algorithm can improve the accuracy and convergence speed of the proposed algorithm. Aiming at the premature convergence of gravitational search algorithm, an adaptive mutation gravitational search algorithm is proposed. The algorithm combines the Metropolis criterion of simulated annealing algorithm (SA) and the hybrid mutation strategy of adaptive mutation probability with the gravity search algorithm. Numerical experiments show that the improved algorithm has better performance.
【学位授予单位】：渤海大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【参考文献】