电子学报 ›› 2022, Vol. 50 ›› Issue (7): 1664-1673.DOI: 10.12263/DZXB.20201305

• 学术论文 • 上一篇    下一篇

量子启发式优化算法的尺度动态调速机制

穆磊1,2, 王鹏1   

  1. 1.西南民族大学计算机科学与工程学院,四川 成都 610041
    2.电子科技大学计算机科学与工程学院,四川 成都 611731
  • 收稿日期:2020-11-18 修回日期:2021-03-01 出版日期:2022-07-25
    • 作者简介:
    • 穆 磊 男,1982年2月出生于河南省三门峡市.2011年在北京理工大学获得工学博士学位,现为西南民族大学讲师.研究方向为群体智能、物联网技术.E-mail: truemoller@126.com
      王 鹏(通讯作者) 男,1975年8月出生于四川省乐山市.现为西南民族大学教授、中科院成都计算所博士生导师.研究方向为智能算法、云计算、并行计算.E-mail: wp002005@163.com
    • 基金资助:
    • 四川省科技厅重点研发项目 (2019YFG0536); 中央高校基本科研业务费 (2020NQN20)

Speed Regulation of Scale Adjustment in Quantum-Inspired Optimization Algorithm

MU Lei1,2, WANG Peng1   

  1. 1.School of Computer Science and Engineering, Southwest Minzu University, Chengdu, Sichuan 610041, China
    2.School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan 611731, China
  • Received:2020-11-18 Revised:2021-03-01 Online:2022-07-25 Published:2022-07-30
    • Supported by:
    • Key Research and Development Program of Science and Technology Department of Sichuan Province (2019YFG0536); Fundamental Research Funds for the Central Universities (2020NQN20)

摘要:

尺度在量子启发谐振子优化算法中起着重要作用,反映了解空间中搜索探针的分辨率.当前研究中以固定速度调整尺度并未合理利用尺度资源.此外,候选解可能会因高斯采样的聚集效应而陷入边界.本文提出了一种在一定程度上反映了适应度利用效率的指标,称为适应度进化利用率.在此基础上,本文提出了一种具有尺度动态调速机制和边界映射反弹策略的量子启发式优化方法.该算法通过与尺度调整因子相关的适应度进化利用率动态调节尺度调整速度,通过2种不同的边界映射反弹策略增加可行解的多样性.将本算法与多种流行优化算法在基准测试函数集上进行对比实验,采用了一种带有动态可接受误差的成功率评估机制保证公平性,实验结果表明该算法具有较强的竞争性.

关键词: 速度调节, 优化算法, 尺度调整, 量子行为

Abstract:

The scale plays an important role in the multiscale quantum-inspired harmonic oscillator algorithm, and it reflects the resolution of searching probes in solution space. The fixed-rate speed of scale adjustment in the current researches leads to unreasonable use of the resources on a specific scale. Besides, candidate solutions may be trapped in the boundary for the aggregation effect of Gaussian sampling. A metric called fitness evolution ratio, which reflects the fitness utilization efficiency to a certain extent, is proposed in this paper. On this basis, a quantum-inspired optimization algorithm is put forward with speed regulation of scale adjustment and boundary mapping rebounding strategy. The algorithm dynamically adjusts the scale adjustment speed through the fitness evolution ratio related to the scale adjustment factor. Besides, it implements two different boundary mapping rebounding strategies to increase the diversity of candidate solutions. Comparison experiments are conducted on the benchmark functions with a variety of compared algorithms. For a fair comparison, an evaluation mechanism of success rate with dynamic acceptable errors is utilized. The results show considerable competitiveness for our scheme.

Key words: speed regulation, optimization algorithm, scale adjustment, quantum behavior

中图分类号: