电子学报 ›› 2021, Vol. 49 ›› Issue (9): 1724-1735.DOI: 10.12263/DZXB.20200593

• 学术论文 • 上一篇    下一篇

基于阶梯式Tent混沌和模拟退火的樽海鞘群算法

周鹏1,2, 董朝轶1,2, 陈晓艳1,2, 齐咏生1,2, 赵肖懿1,2, 王启来1,2   

  1. 1.内蒙古工业大学,内蒙古 呼和浩特 010080
    2.内蒙古机电控制重点实验室(内蒙古工业大学),内蒙古 呼和浩特 010051
  • 收稿日期:2020-06-18 修回日期:2021-01-23 出版日期:2021-10-21
    • 作者简介:
    • 周 鹏 男,1991年2月生于山东省菏泽市. 研究生.主要研究方向为智能优化算法、系统辨识与控制方法.E‑mail: zhoupeng_hz@163.com
      董朝轶(通信作者) 男,1976年7月生于内蒙古自治区包头市.现为内蒙古工业大学教授、硕士生导师,主要研究方向为地面移动机器人的动态建模、导航、复杂生物网络建模与网络结构辨识.E‑mail:dongchaoyi@hotmail.com
    • 基金资助:
    • 国家自然科学基金 (61863029); 内蒙古科技成果转化项目 (CGZH2018129); 内蒙古自治区科技计划项目申报书 (关键技术攻关计划项目); 内蒙古自然科学基金 (2020MS06020)

A Salp Swarm Algorithm Based on Stepped Tent Chaos and Simulated Annealing

ZHOU Peng1,2, DONG Chao-yi1,2, CHEN Xiao-yan1,2, QI Yong-sheng1,2, ZHAO Xiao-yi1,2, WANG Qi-lai1,2   

  1. 1.Inner Mongolia University of Technology, Hohhot, Inner Mongolia 010080, China
    2.Inner Mongolia Electromechanical Control Laboratory, Hohhot, Inner Mongolia 010051, China
  • Received:2020-06-18 Revised:2021-01-23 Online:2021-10-21 Published:2021-09-25
    • Supported by:
    • National Natural Science Foundation of China (61863029); Program of Scientific and Technological Achievement Transformation of Inner Mongolia Autonomous Region (CGZH2018129); Application for Science and Technology Project of Inner Mongolia Autonomous Region; Natural Science Foundation of Inner Mongolia Autonomous Region, China (2020MS06020)

摘要:

针对樽海鞘群算法寻优迭代过程中存在容易陷入局部最优、收敛速度慢的问题,提出一种改进的樽海鞘群算法.引入Tent混沌映射初始化种群来提高算法迭代前期的收敛速度,通过惯性权值“阶梯式”调整策略来更好地兼顾算法全局探索能力和局部开发能力,通过模拟退火增强樽海鞘群算法迭代后期跳出局部最优解的能力,以基准测试函数和磁导航自动导引车模糊控制器参数寻优问题为例测试了算法性能.仿真结果表明,对于单峰和多峰测试函数,改进后的樽海鞘群算法具有更快的收敛速度和更强的全局寻优能力.相比较标准樽海鞘群算法的参数调节法,改进后的樽海鞘群算法所设计的磁导航自动导引车模糊控制器对磁偏差值控制性能更为优化,在控制器设计方面具有潜在的应用价值.

关键词: 樽海鞘群算法, Tent混沌映射, 阶梯式权值, 模拟退火, 系统辨识, 自动导引车

Abstract:

To solve the problems of local optimization and slow convergence in the process of optimization and iteration, the paper proposed an improved optimization algorithm for a salp swarm, i.e, step?by?step tent chaos simulated annealing salp swarm algorithm (STSA?SSA). Firstly, an initial population of Tent chaotic map was introduced to enhance the algorithm convergence at the early stage of iteration, and a step adjustment strategy of inertia weights was employed to improve the global and local exploring ability of the STSA?SSA. Then, the ability of escaping local optimal solutions of the STSA?SSA at the later stage of iterations was increased by a simulated annealing policy. Finally, the performance of the STSA?SSA was tested in the processes of optimizing the parameters of benchmark functions and a fuzzy controller a magnetic navigation automated guided vehicle (AGV). The results show that for the single peak and multi peak test functions, the STSA?SSA has faster convergence speed and stronger global optimization ability. Compared with SSA, the fuzzy controller of the magnetic navigation AGV designed by STSA?SSA is more optimized. Therefore, the STSA?SSA has a potential engineering application value in controller designs.

Key words: salp swarm algorithm, tent chaotic mapping, step weight, simulated annealing, system identification, AGV

中图分类号: