期刊
IEEE ACCESS
卷 8, 期 -, 页码 21909-21916出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2020.2968980
关键词
Electromagnetic application; optimization design; particle swarm optimizer; quantum mechanics; mutation mechanism
资金
- National Natural Science Foundation of China [61801008]
- National Key Research and Development Program of China [2018YFB0803600]
- Beijing Natural Science Foundation [L172049]
- Scienti~c Research Common Program of Beijing Municipal Commission of Education [KM201910005025]
Quantum inspired particle swarm optimization (QPSO) stimulated by perceptions from particle swarm optimization and quantum mechanics is a stochastic optimization method. Although, it has shown good performance in finding the optimal solution to many electromagnetic problems. However, sometimes it falls to local optima when dealing with hard optimization problems. Thus, to preserve a good balance between local and global searches to avoid premature convergence in quantum particle swarm optimization, this paper proposed three enhancements to the original QPSO method, the proposed method is called modified quantum particle swarm optimization (MQPSO) algorithm. Firstly, a novel selection technique is introduced that will choose the best particle among the population within the search domain to achieve a high-performance exploration. Secondly, a new mutation method is used to preserve the easiness of available QPSOs. Also, a dynamic parameter strategy is proposed for further facilitating the algorithm and tradeoff between exploration and exploitation searches. The experimental results obtained by solving standard benchmark functions and an electromagnetic design problem which is the superconducting magnetic energy storage (SMES) system available in both three parameters and eight parameters problems are reported to showcase the usefulness of the proposed approach.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据