Scholarlink

<p>Although quantum-behaved particle swarm optimization (QPSO) possesses a user-friendly operation and remarkable performance for various applications, it is apt to prematurity while solving high-dimensional multimodal problems. For enhancing the global searching efficiency of QPSO, an optimized QPSO based on hybrid strategy (HSQPSO) is presented, and its improvement approaches involve the following three aspects. Firstly, subpopulation initialization strategy is applied to help the algorithm avoid premature convergence. Next, a coevolutionary pattern based on information interaction is developed to improve the global optimization capability of QPSO. Subsequently, an adjustment operator for the global optimal location is implemented to assist QPSO in coming out from local optimum. The results of the comparative experiments show that HSQPSO outperforms the selected algorithms in terms of optimization accuracy and significance in more than 60% of the test functions, and it takes less time to run in more than 90% of the test functions. Further, four engineering application issues from the real world are employed to testify the enhancement of HSQPSO. Compared results obtained by HSQPSO and other optimization approaches indicate that HSQPSO can display superior global search capability and highly adaptability in handling different multimodal cases.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantum-behaved particle swarm optimizer driven by hybrid strategy and its applications in engineering optimization

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Guang He,
  • Xiao-li Lu

摘要

Although quantum-behaved particle swarm optimization (QPSO) possesses a user-friendly operation and remarkable performance for various applications, it is apt to prematurity while solving high-dimensional multimodal problems. For enhancing the global searching efficiency of QPSO, an optimized QPSO based on hybrid strategy (HSQPSO) is presented, and its improvement approaches involve the following three aspects. Firstly, subpopulation initialization strategy is applied to help the algorithm avoid premature convergence. Next, a coevolutionary pattern based on information interaction is developed to improve the global optimization capability of QPSO. Subsequently, an adjustment operator for the global optimal location is implemented to assist QPSO in coming out from local optimum. The results of the comparative experiments show that HSQPSO outperforms the selected algorithms in terms of optimization accuracy and significance in more than 60% of the test functions, and it takes less time to run in more than 90% of the test functions. Further, four engineering application issues from the real world are employed to testify the enhancement of HSQPSO. Compared results obtained by HSQPSO and other optimization approaches indicate that HSQPSO can display superior global search capability and highly adaptability in handling different multimodal cases.