摘要
针对现有球形机器人存在环境感知能力较弱、运动控制精度不高的问题,提出一种具有立体视觉的新型球形机器人Visionbot。介绍Visionbot机器人的机械结构,采用具有非完整约束的拉格朗日方程建立该球形机器人的动力学模型。针对状态变量过多难以用于控制系统设计的问题,对球形机器人的动力学模型进行合理简化。利用模型解耦、线性化以及状态反馈等方法,建立该球形机器人的运动控制模型,包括内环力矩控制器和外环路径跟踪控制器。采用延时环节对视觉反馈通道的时延进行建模,建立具有视觉反馈的球形机器人运动控制模型。通过运动仿真试验,研究机器人重摆质量、视觉反馈延时对其运动控制性能的影响,并验证系统设计的可行性。研究结果为Visionbot机器人的优化设计和样机研制提供理论基础。
Because existing ball-shaped robots have many drawbacks, such as weak capability of environmental perception and low precision of motion control, a novel ball-shaped robot with stereo vision, referred to as Visionbot, is proposed. After introduction of mechanical structure of the Visionbot, dynamic model of the robot is developed by using the constrained general Lagrangian formulation, as well as model simplification under reasonable assumptions. Based on a linearization approach, dynamic model of the robot is decomposed, and motion controller is designed, including two inner torque controllers and an outer path tracking controller. A motion control algorithm with delayed visual feedback is designed for the robot. Effects of pendulum mass and delayed visual feedback on performance of the control algorithm is studied by simulated experiments. Experimental resuRs not only validate the feasibility of system design, but also can be used to do the optimum design and prototype development of the robot.
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2013年第11期8-15,共8页
Journal of Mechanical Engineering
基金
国家自然科学基金(51175048)
中央高校基本科研业务费专项基金(2009RC0601)
高等学校科技创新工程重大培育(708011)资助项目
关键词
球形机器人
运动控制
立体视觉
路径跟踪
Ball-shaped robot Motion control Stereo vision Path tracking