群体智能优化算法-《郑州大学学报(工学版)》

文章信息/Info

作者:: 程适; 王锐; 伍国华; 郭一楠; 马连博; 史玉回; 1. 陕西师范大学计算机科学学院;2. 国防科技大学系统工程学院;3. 中南大学交通运输工程学院;4. 中国矿业大学信息与控制工程学院;5. 东北大学软件学院;6. 南方科技大学计算机科学与工程系

Author(s):: Cheng Shi 1; Wang Rui 2; Wu Guohua 3; Guo Yinan 4; Malembo 5; Shi Yuhui 6; 1. School of Computer Science, Shaanxi Normal University; 2. School of Systems Engineering, National University of Defense Technology; 3. School of Transportation Engineering, Central South University; 4. School of Information and Control Engineering, China University of Mining and Technology; 5. School of Software, Northeastern University; 6. Department of Computer Science and Engineering, Southern University of Science and Technology

Keywords:: Swarm intelligence; optimization algorithms; intelligent computing

摘要:: 群体智能( swarm intelligence) 的核心思想就是若干个简单个体构成一个群体，通过合作、竞争、交互与学习等机制表现出高级和复杂的功能，在缺少局部信息和模型的情况下，仍能够完成复杂问题的求解．其求解过程为对求解变量进行随机初始化，经过迭代求解，计算目标函数的输出值．群体智能优化算法不依赖于梯度信息，对待求解问题无连续、可导等要求，使得该类算法既适应连续型数值优化，也适应离散型组合优化．同时，群体智能优化算法潜在的并行性和分布式特点使其在处理大数据时具备显著优势．

Abstract:: The core idea of swarmintelligence (swarmintelligence) is that several simple individuals form a group, through cooperation, competition, interaction and learning mechanisms to show advanced and complex functions, in the absence of local information and models, still able to complete the solution of complex problems.The solution process is to initialize the variable randomly, and calculate the output value of the objective function after iterative solution.Swarm intelligent optimization algorithm is not dependent on gradient information, and it is not continuous and derivable to solve problems, which makes it suitable for both continuous numerical optimization and discrete combinational optimization.At the same time, the potential parallelism and distributed characteristics of swarm intelligence optimization algorithm make it have significant advantages in dealing with big data.