Journal of Frontiers of Computer Science and Technology ›› 2020, Vol. 14 ›› Issue (10): 1644-1655.DOI: 10.3778/j.issn.1673-9418.1912014

Previous Articles     Next Articles

Convergence Analysis and Performance Comparison of Cuckoo Search Algorithm

LIU Xiaodong, SUN Lijun, CHEN Tianfei   

  1. 1. Key Laboratory of Grain Information Processing and Control of Ministry of Education, Henan University of Technology, Zhengzhou 450001, China
    2. Zhengzhou Key Laboratory of Machine Perception and Intelligent System, Henan University of Technology, Zhengzhou 450001, China
    3. College of Electrical Engineering, Henan University of Technology, Zhengzhou 450001, China
    4. College of Information Science and Engineering, Henan University of Technology, Zhengzhou 450001, China
  • Online:2020-10-01 Published:2020-10-12

布谷鸟算法的收敛性分析及性能比较

刘晓东孙丽君陈天飞   

  1. 1. 河南工业大学 粮食信息处理与控制教育部重点实验室,郑州 450001
    2. 河南工业大学 郑州市机器感知与智能系统重点实验室,郑州 450001
    3. 河南工业大学 电气工程学院,郑州 450001
    4. 河南工业大学 信息科学与工程学院,郑州 450001

Abstract:

Swarm intelligence algorithm is an emerging optimization method, which connects simple individuals through teamwork and organization to generate swarm intelligence and is used to solve practical problems. So far, there are many kinds of swarm intelligence algorithms. Cuckoo search (CS) is one of the typical swarm intelligence algorithms. It has the characteristics of simple implementation and high efficiency. In this paper, based on the principle of CS algorithm, a Markov chain model is constructed, its properties are analyzed, and the global convergence criterion is combined to prove the global convergence of CS algorithm. The experiment is simulated in two aspects: firstly, in the case of solving the same problem, this paper analyzes the complexity of 5 algorithms; secondly, 18 standard test functions are adopted to perform statistics on the 5 algorithms respectively. Under low and high dimension, the accuracy, convergence speed and stability of the 5 kinds of algorithms are compared. The experimental results show that cuckoo algorithm is of low complexity, and compared with other algorithms, it has high accuracy and good stability.

Key words: swarm intelligence algorithm, cuckoo search (CS), Markov chain, global convergence, accuracy

摘要:

群体智能算法是一类新兴的优化方法,它通过团队的协作和组织将简单的个体联系起来产生群体智慧,并用于解决实际问题。迄今为止,群体智能算法种类繁多,布谷鸟算法(CS)是典型的群体智能算法之一,它具有实现简单、效率较高等特点。以标准CS算法原理为基础构建Markov链模型,分析其性质,结合全局收敛准则,证明CS算法的全局收敛性。实验在两方面进行仿真:一方面,在解决相同问题的情况下,分析5种算法的复杂度;另一方面,选取18个标准测试函数分别对5种算法进行数据统计,在低维度和高维度下对比了5种算法运行的精确度、收敛速度和稳定性。实验结果表明:布谷鸟算法复杂度较低,与其他算法相比,其精度高,稳定性好。

关键词: 群体智能算法, 布谷鸟算法(CS), Markov链, 全局收敛性, 精确度