基于运动描述语言的机械臂轨迹生成及仿真
|
摘要
传统的机械臂轨迹生成与控制方法本质上是一种离散采样方法,主控制器以固定的时间间隔采样关节的轨迹参考函数,并将采样点的位置、速度等信息作为参考控制信号发送给关节控制器。在这种方法中,轨迹精度依赖于采样点的数量,提高精度就要增加采样点,但是这将增加数据传输量,总线系统的速率和带宽也需要相应提高。针对传统方法的特点,提出基于运动描述语言(MDL)的机械臂轨迹生成与控制方法。上述方法把连续的轨迹函数映射到运动基元,用运动基元的组合来描述轨迹函数,总线系统只需实时传输少量参数化的运动基元,就可以在关节端重构参考轨迹。上述方法摆脱了离散采样的方式,可以在不提高总线速率的前提下提高机械臂的轨迹精度,还通过仿真软件验证了上述方法的有效性。
Traditional trajectory generation and control method is essentially a discrete sampling method. The joint's trajectory function is sampled at a fixed time interval in the core controller, then the position, velocity and so on are transformed to the joint controller. In this method, the accuracy of trajectory depends on the number of samples, which means the more samples the more accuracy. However the bandwidth and rate of communication bus are limited, the number of samples cannot be increased arbitrarily. In view of the feature of the traditional method, this paper proposes a method of manipulator trajectory generation and control based on motion description language(MDL). We mapped continuous trajectory functions to some atoms, and described these functions by the combination of atoms. A few of parameterized atoms were transmitted by the communication bus, and the reference function can be reconstructed in the joint controller. This method gets rid of the sampling way. The accuracy of trajectory can be improved on the premise of the same communication rate. This article also verifies the validity of the method via simulation software.
Traditional trajectory generation and control method is essentially a discrete sampling method. The joint's trajectory function is sampled at a fixed time interval in the core controller, then the position, velocity and so on are transformed to the joint controller. In this method, the accuracy of trajectory depends on the number of samples, which means the more samples the more accuracy. However the bandwidth and rate of communication bus are limited, the number of samples cannot be increased arbitrarily. In view of the feature of the traditional method, this paper proposes a method of manipulator trajectory generation and control based on motion description language(MDL). We mapped continuous trajectory functions to some atoms, and described these functions by the combination of atoms. A few of parameterized atoms were transmitted by the communication bus, and the reference function can be reconstructed in the joint controller. This method gets rid of the sampling way. The accuracy of trajectory can be improved on the premise of the same communication rate. This article also verifies the validity of the method via simulation software.
引文
[1] R Brockett.Formal languages for motion description and map making[J].Robotics,1990,41:181-191.
[2] R W Brockett.On the computer control of movement[C].Robotics and Automation,1988.Proceedings,1988 IEEE International Conference on,1988:534-540.
[3] Brockett R W.Hybrid models for motion control systems[M].Springer,1993.
[4] V Manikonda,P S Krishnaprasad,J Hendler.A motion description language and a hybrid architecture for motion planning with nonholonomic robots[J].Proceedings of 1995 Ieee International Conference on Robotics and Automation,Vols 1-3,1995 2021-2028.
[5] H Li.Choreographing Dynamical Systems[D].Harvard University,2004.
[6] 化建宁.基于网络的机器人遥操作系统:运动描述语言方法[D].中国科学院沈阳自动化研究所,2007.
[7] 化建宁,等.基于MDL的机器人网络遥操作系统控制方法[J].机器人,2013,35(5):615-622.
[8] 化建宁,等.基于运动描述语言的轮式移动机器人控制[J].机器人,2006,28(3):316-320.
[9] P J Martin.Motion description languages:from specification to execution[D].Georgia Institute of Technology,2010.
[10] Gargas Iii E F.Generation and use of a discrete robotic controls alphabet for high-level tasks[D].Georgia Institute of Technology,2012.
[11] 刘维惠,陈殿生,张立志.人机协作下的机械臂轨迹生成与修正方法[J].机器人,2016,38(4):504-512.
[12] Craig J J,贠超.机器人学导论[M].北京:机械工业出版社,2006.
[13] Biagiotti L,Melchiorri C.Trajectory Planning for Automatic Machines and Robots[M].Springer Science & Business Media,2008.
(1)仿真代码及录像https://github.com/liuzhaoming5954/MDL_paper
[2] R W Brockett.On the computer control of movement[C].Robotics and Automation,1988.Proceedings,1988 IEEE International Conference on,1988:534-540.
[3] Brockett R W.Hybrid models for motion control systems[M].Springer,1993.
[4] V Manikonda,P S Krishnaprasad,J Hendler.A motion description language and a hybrid architecture for motion planning with nonholonomic robots[J].Proceedings of 1995 Ieee International Conference on Robotics and Automation,Vols 1-3,1995 2021-2028.
[5] H Li.Choreographing Dynamical Systems[D].Harvard University,2004.
[6] 化建宁.基于网络的机器人遥操作系统:运动描述语言方法[D].中国科学院沈阳自动化研究所,2007.
[7] 化建宁,等.基于MDL的机器人网络遥操作系统控制方法[J].机器人,2013,35(5):615-622.
[8] 化建宁,等.基于运动描述语言的轮式移动机器人控制[J].机器人,2006,28(3):316-320.
[9] P J Martin.Motion description languages:from specification to execution[D].Georgia Institute of Technology,2010.
[10] Gargas Iii E F.Generation and use of a discrete robotic controls alphabet for high-level tasks[D].Georgia Institute of Technology,2012.
[11] 刘维惠,陈殿生,张立志.人机协作下的机械臂轨迹生成与修正方法[J].机器人,2016,38(4):504-512.
[12] Craig J J,贠超.机器人学导论[M].北京:机械工业出版社,2006.
[13] Biagiotti L,Melchiorri C.Trajectory Planning for Automatic Machines and Robots[M].Springer Science & Business Media,2008.
(1)仿真代码及录像https://github.com/liuzhaoming5954/MDL_paper