仿生并联眼动机构多目标优化设计
仿生并联眼动机构多目标优化设计仿生并联眼动机构多目标优化设计
程元皓, 王孙安, 于德弘
西安交通大学机械工程学院, 陕西 西安 710049
Multi-objective Optimization Design of a Bionic Parallel Oculogyric Mechanism
CHENG Yuanhao, WANG Sun'an, YU Dehong
School of Mechanical Engineering, Xi'an Jiaotong University, Xi'an 710049, China
全文: PDF (493 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)
摘要 针对低空空域视觉监控系统需同时保证大范围搜索和对目标持续平滑跟踪的要求，依据仿生原理，设计了一种模拟人类眼外肌功能的仿生并联眼动机构构型．为克服当前并联眼动机构多采用单目标优化方法而无法保障动平台倾角及运动精度、机构灵巧度、运动传递性能等运动性能同时取优的问题，在安装空间、驱动件尺寸、铰链偏角等约束条件下，采用改进非支配排序遗传算法（NSGA-Ⅱ）优化机构的结构参数．根据优化得到的结构参数搭建了实验样机．实际测量结果表明，所设计的机构能够在动平台最大倾角、综合运动性能方面取得较为理想的效果，动平台倾角可达55.94°，精度可达0.01°，优于人类视觉系统及相关研究，满足低空空域监控需求．
关键词 ： 仿生原理,眼外肌,并联眼动机构,多目标优化,改进非支配排序遗传算法
Abstract：To ensure requirements of wide search range and smooth and continuous target-tracking in the low-altitude airspace visual monitoring system, a bionic parallel oculogyric mechanism is designed to simulate the function of human extraocular muscles according to the bionic principle. Most of the existing parallel oculogyric mechanisms adopt the single-objective optimization, which can't simultaneously optimize the dip angle of the moving platform and the motion performances including the motion precision, the mechanism dexterity and the motion transmission performance. In order to overcome the problem, non-dominated sorting-based genetic algorithm-Ⅱ (NSGA-Ⅱ) is used to optimize the structural parameters of the mechanism under the constraints of the installation, driver size and hinge angle. An experimental prototype is developed according to the optimized structural parameters. The actual measurement results show that the mechanism designed can achieve satisfactory results in terms of the maximum dip angle and the integrated motion performance of the moving platform, and the dip angle of 55.94° and the precision of 0.01° are realized by the moving platform, which are better than the human vision system as well as related research, and also meet the needs of low-altitude airspace monitoring.
Key words： bionic principle extraocular muscle parallel oculogyric mechanism multi-objective optimization NSGA-II (non-dominated sorting-based genetic algorithm-II)
收稿日期: 2016-10-08 出版日期: 2017-02-15
1:TP242.3
基金资助:国家自然科学基金（51375368）
通讯作者: 王孙安,sawang@mail.xjtu.edu.cn E-mail: sawang@mail.xjtu.edu.cn
作者简介: 程元皓(1987-),男,博士生.研究领域:机器人仿生技术,机器人控制.
王孙安(1957-),男,博士,教授.研究领域:机器人仿生技术,机器人控制.
引用本文:
程元皓, 王孙安, 于德弘. 仿生并联眼动机构多目标优化设计. 机器人, 2017, 39(2): 176-181.
CHENG Yuanhao, WANG Sun'an, YU Dehong. Multi-objective Optimization Design of a Bionic Parallel Oculogyric Mechanism. ROBOT, 2017, 39(2): 176-181.
链接本文:
http://robot.sia.cn/CN/10.13973/j.cnki.robot.2017.0176 或 http://robot.sia.cn/CN/Y2017/V39/I2/176
页:
[1]