模糊优化结合智能干扰区域划分的微装配控制
|
摘要
针对传统机器人零件微装配算法不能有效解决干扰区域划分问题,提出一种模糊优化结合机器学习的智能干扰区域划分算法。首先,根据工作区内零件与目标孔之间的位置关系设计模糊协调器和特殊的规则库,避免激活干扰状态。然后,利用提出的智能区域划分算法合并所有相邻区域。最后,通过决策制定,选择区域中最低模糊熵的四元组控制值作为执行分配任务的最终控制值。实验结果表明,该算法可将69个子区域合并降至10个子区域,相比其他的较为先进的装配算法,算法更加灵活,显著提高了任务效率。
As the traditional algorithm for the micro assemblage of the robot parts cannot effectively solve the problem of jamming region division, an intelligent jamming region division algorithm based on fuzzy optimization combined with intelligent machine learning is proposed. First, the fuzzy coordinator and special rule base is designed by the position relationship between parts in the work space and the target holes, avoiding the activation state interference. Then, the adjacent areas are merged with the proposed region division algorithm. Finally, by the decision-making, the control value of the four-tuple whose fuzzy entropy is the lowest in the regions is chosen as the final control value to assign tasks. Experimental results show that 69 sub-regions can be merged into only 10 sub-regions by the proposed algorithm. And the 10 sub-regions can show all the possible assemblage states. Compared with other advanced assembly algorithms, the proposed algorithm is more flexible, and it has significantly improved the efficiency of the assemblage task.
As the traditional algorithm for the micro assemblage of the robot parts cannot effectively solve the problem of jamming region division, an intelligent jamming region division algorithm based on fuzzy optimization combined with intelligent machine learning is proposed. First, the fuzzy coordinator and special rule base is designed by the position relationship between parts in the work space and the target holes, avoiding the activation state interference. Then, the adjacent areas are merged with the proposed region division algorithm. Finally, by the decision-making, the control value of the four-tuple whose fuzzy entropy is the lowest in the regions is chosen as the final control value to assign tasks. Experimental results show that 69 sub-regions can be merged into only 10 sub-regions by the proposed algorithm. And the 10 sub-regions can show all the possible assemblage states. Compared with other advanced assembly algorithms, the proposed algorithm is more flexible, and it has significantly improved the efficiency of the assemblage task.
引文
[1]蔡锦达,张剑皓,秦绪祥.六轴工业机器人的参数辨识方法[J].控制工程,2013,20(5):805-808.Cai J D,Zhang J H,Qin X X.Parameter identification method of6-axis industrial robot[J].Control Engineering of China,2013,20(5):805-808.
[2]沈飞,徐德,唐永建,等.微操作/微装配中微力觉的测量与控制技术研究现状综述[J].自动化学报,2014,37(5):1352-1358.Sheng F,Xu D,Tang Y J,et al.Review of measuring and control technology of microforce in micromanipulation and microassembly[J].Acta Automatica Sinica,2014,37(5):1352-1358.
[3]Chang H C,Liang G S,Chu C W,et al.Prioritizing service attributes for improvement using fuzzy zone of tolerance[J].International Journal of Innovative Computing,Information and Control,2012,8(1):75-89.
[4]Cecil J,Jones J.VREM:An advanced virtual environment for micro assembly[J].The International Journal of Advanced Manufacturing Technology,2014,72(1-4):47-56.
[5]Jain R K,Majumder S,Bano A,et al.Stability Analysis of Piezoelectric Actuator based Micro Gripper for Robotic Micro Assembly[C]//Proceedings of Conference on Advances In Robotics.ACM,2013:1-6.
[6]郝永平,王永杰,董福禄,等.平板类微小零件装配控制策略与软件架构研究[J].机械工程学报,2015,51(4):193-198.Hao Y P,Wang Y J,Dong F L,et al.Study of control strategy and software architecture based on flat type small parts assembly[J].Journal of Mechanical Engineering,2015,51(4):193-198.
[7]张娟,徐德,张正涛等.基于多路显微视觉的微零件自动对准策略[J].机器人,2014,36(1):69-75.Zhang J,Xu D,Zhang Z T,et al.An automatic alignment strategy of micro parts based on microscope vision systems[J].Robot,2014,36(1):69-75.
[8]Qiu H,Zhang H.Fuzzy SLIQ decision tree based on classification sensitivity[J].International Journal of Modern Education and Computer Science(IJMECS),2011,3(5):108-115.
[9]史玉珍,吕琼帅.基于进化模糊规则的Web新闻文本挖掘与分类方法[J].湘潭大学自然科学学报,2016,38(2):99-103.SHI Y Z,LV Q S.Web News Text Mining and Classification Method Based on Evolving Fuzzy Rule[J].Natural Science Journal of Xiangtan University,2016,38(2):99-103.
[10]朱其新,张正,杨辉等.基于无传感器的PMSM电流控制策略的研究[J].控制工程,2014,21(4):1245-1251.Zhu Q X,Zhang Y,Yang H,et al.Research on sensorless PMSM current control strategies[J].Control Engineering of China,2014,21(4):1245-1251.
[11]李江昊.基于毫米级移动微机器人的微装配系统运动控制与路径规划研究[D].上海:上海交通大学,2009.Li J H.Research on motion control and path planning of micro assembly system based on millimeter level mobile micro robot[D].Shanghai:Shanghai Jiaotong University,2009.
[12]Wang Y,Li L,Cui G,et al.Ontogenesis from embryo to juvenile and salinity tolerance of Japanese devil stinger Inimicus japonicus during early life stage[J].Springerplus,2013,2(1):45-51.
[13]Aliev R,Pedrycz W,Fazlollahi B,et al.Fuzzy logic-based generalized decision theory with imperfect information[J].Information Sciences,2012,18(9):18-42.
[14]Essen M V,Hirvonen J,Kuikka S,et al.Robotic software frameworks and software component models in the development of automated handling of individual natural fibers[J].Journal of Micro-Bio Robotics,2014,9(1-2):29-45.
[2]沈飞,徐德,唐永建,等.微操作/微装配中微力觉的测量与控制技术研究现状综述[J].自动化学报,2014,37(5):1352-1358.Sheng F,Xu D,Tang Y J,et al.Review of measuring and control technology of microforce in micromanipulation and microassembly[J].Acta Automatica Sinica,2014,37(5):1352-1358.
[3]Chang H C,Liang G S,Chu C W,et al.Prioritizing service attributes for improvement using fuzzy zone of tolerance[J].International Journal of Innovative Computing,Information and Control,2012,8(1):75-89.
[4]Cecil J,Jones J.VREM:An advanced virtual environment for micro assembly[J].The International Journal of Advanced Manufacturing Technology,2014,72(1-4):47-56.
[5]Jain R K,Majumder S,Bano A,et al.Stability Analysis of Piezoelectric Actuator based Micro Gripper for Robotic Micro Assembly[C]//Proceedings of Conference on Advances In Robotics.ACM,2013:1-6.
[6]郝永平,王永杰,董福禄,等.平板类微小零件装配控制策略与软件架构研究[J].机械工程学报,2015,51(4):193-198.Hao Y P,Wang Y J,Dong F L,et al.Study of control strategy and software architecture based on flat type small parts assembly[J].Journal of Mechanical Engineering,2015,51(4):193-198.
[7]张娟,徐德,张正涛等.基于多路显微视觉的微零件自动对准策略[J].机器人,2014,36(1):69-75.Zhang J,Xu D,Zhang Z T,et al.An automatic alignment strategy of micro parts based on microscope vision systems[J].Robot,2014,36(1):69-75.
[8]Qiu H,Zhang H.Fuzzy SLIQ decision tree based on classification sensitivity[J].International Journal of Modern Education and Computer Science(IJMECS),2011,3(5):108-115.
[9]史玉珍,吕琼帅.基于进化模糊规则的Web新闻文本挖掘与分类方法[J].湘潭大学自然科学学报,2016,38(2):99-103.SHI Y Z,LV Q S.Web News Text Mining and Classification Method Based on Evolving Fuzzy Rule[J].Natural Science Journal of Xiangtan University,2016,38(2):99-103.
[10]朱其新,张正,杨辉等.基于无传感器的PMSM电流控制策略的研究[J].控制工程,2014,21(4):1245-1251.Zhu Q X,Zhang Y,Yang H,et al.Research on sensorless PMSM current control strategies[J].Control Engineering of China,2014,21(4):1245-1251.
[11]李江昊.基于毫米级移动微机器人的微装配系统运动控制与路径规划研究[D].上海:上海交通大学,2009.Li J H.Research on motion control and path planning of micro assembly system based on millimeter level mobile micro robot[D].Shanghai:Shanghai Jiaotong University,2009.
[12]Wang Y,Li L,Cui G,et al.Ontogenesis from embryo to juvenile and salinity tolerance of Japanese devil stinger Inimicus japonicus during early life stage[J].Springerplus,2013,2(1):45-51.
[13]Aliev R,Pedrycz W,Fazlollahi B,et al.Fuzzy logic-based generalized decision theory with imperfect information[J].Information Sciences,2012,18(9):18-42.
[14]Essen M V,Hirvonen J,Kuikka S,et al.Robotic software frameworks and software component models in the development of automated handling of individual natural fibers[J].Journal of Micro-Bio Robotics,2014,9(1-2):29-45.