基于改进粒子群算法的分数阶系统参数辨识(英文)
发布时间:2021-05-01 03:27
为了更好地辨识分数阶系统的参数,提出了一种基于Tent映射的改进粒子群算法(MPSO).采用8个经典测试函数对MPSO算法的性能进行了测试,并与自适应时变加速器算法(ACPSO)、改进的被动聚集粒子群算法(IPSO)以及遗传算法(GA)进行对比,验证了所提算法的有效性.在已知模型结构和未知模型结构的基础上,利用所提算法对2种典型分数阶模型进行参数辨识.参数辨识结果表明,应用位置信息的平均值有利于充分共享个体间的信息,从而能够加快全局搜索速度;Tent映射具有的均匀性和遍历性能够防止位置信息中极值的产生,避免算法陷入局部最优.MPSO算法收敛速度快、精度高,是一种有效且实用的方法.
【文章来源】:Journal of Southeast University(English Edition). 2018,34(01)EI
【文章页数】:9 页
【文章目录】:
1 Theory
1.1 Definition of fractional-order derivatives and integrals
1.2 Fractional-order systems
1.3 PSO variants
1.3.1 ACPSO algorithm
1.3.2 IPSO algorithm
2 Proposed MPSO Algorithm
2.1 Modified Tent mapping
2.2 MPSO algorithm
2.3 Performance evaluation
2.3.1 Classical test functions
2.3.2 Parameter analysis
2.3.3 Evaluation results
3 Simulations
3.1 Identification of known fractional-order model structure
3.2 Identification for unknown fractional-order mod-el structure
4 Conclusions
本文编号:3170053
【文章来源】:Journal of Southeast University(English Edition). 2018,34(01)EI
【文章页数】:9 页
【文章目录】:
1 Theory
1.1 Definition of fractional-order derivatives and integrals
1.2 Fractional-order systems
1.3 PSO variants
1.3.1 ACPSO algorithm
1.3.2 IPSO algorithm
2 Proposed MPSO Algorithm
2.1 Modified Tent mapping
2.2 MPSO algorithm
2.3 Performance evaluation
2.3.1 Classical test functions
2.3.2 Parameter analysis
2.3.3 Evaluation results
3 Simulations
3.1 Identification of known fractional-order model structure
3.2 Identification for unknown fractional-order mod-el structure
4 Conclusions
本文编号:3170053
本文链接:https://www.wllwen.com/projectlw/xtxlw/3170053.html