摘要
粒子群优化算法具有参数设置少,搜索速度快,易于工程实现的优点,广泛应用于神经网络训练程应用领域和随机优化类型等问题的求解。本文通过改变维度,个体数量,迭代次数三个参数,结合4个标准测试函数进行优化实验。验证结果表明,维度,个体数量,迭代次数对粒子群优化算法的优化性能影响较大。
Article swarm optimization(pso)has the advantages of less parameter setting,fast search speed and easy engineering implementation.It is widely used in neural network training engineering application field and the solution of random optimization type.In this paper,three parameters are changed:dimension,number of individuals and number of iterations,and four standard test functions are combined to carry out optimization experiments.The results show that the dimensions,number of individuals and number of iterations have great influence on the optimization performance of pso.
作者
杨俊胜
沈航驰
葛鹏
代永强
YANG Jun-sheng;SHEN Hang-chi;GE Peng;DAI Yong-qiang(College of information science and technology,gansu agricultural university)
出处
《软件》
2020年第5期228-230,共3页
Software
关键词
维度
个体数量
迭代次数
标准测试函数
Dimensions
Number of individuals
Number of iterations
Standard test functions
