论文部分内容阅读
为在环境发生变化后跟踪最优解的变化,提出一种自组织单变量边缘分布算法(SOUMDA)来求解动态优化问题.自组织策略包含扩散和惯性速度模型,扩散模型利用当前环境的局部信息使群体向外扩散,惯性速度模型利用最优解的历史信息进行预测.将自组织策略与单变量边缘分布算法(UMDA)结合,使得算法在环境变化后自适应地增加种群多样性,提高算法适应能力,快速跟踪最优解.利用动态sphere函数对所提出的算法进行测试,并与iUMDA和MUMDA算法进行比较,结果表明所设计的算法能快速适应环境的变化,跟踪最优解.
In order to track the change of the optimal solution after the environment changes, a self-organizing single-variable edge distribution algorithm (SOUMDA) is proposed to solve the dynamic optimization problem.The self-organizing strategy includes the diffusion and inertial velocity models, which use the local information of the current environment Which makes the population spread outwards and the inertial velocity model uses the historical information of the optimal solution to predict.The combination of the self-organizing strategy and the univariate edge distribution algorithm (UMDA) makes the algorithm adaptively increase the population diversity after the environment changes and improve the algorithm Adaptability and fast tracking of the optimal solution.The proposed algorithm is tested with the dynamic sphere function and compared with the iUMDA and MUMDA algorithms. The results show that the proposed algorithm can quickly adapt to environmental changes and track the optimal solution.