On using generalized velocity components for manipulator dynamics and control

In this paper modified first-order decoupled equations of motion for rigid serial manipulators are presented. Motivated the results obtained by Loduha and Ravani [Loduha, T.A., Ravani, B., 1995. On first-order decoupling of equations of motion for constrained dynamical systems. Transactions of the ASME Journal of Applied Mechanics 62, 216.] slightly different inertial quasi-velocities are proposed. Instead of generalized velocity components (GVC) one useful form of GVC’s is introduced here. It is shown that the modified quantities (called here modified inertial generalized velocity components—MIGVC) lead to some interesting properties which give different look at manipulator dynamics. Some properties arising from MIGVC are also discussed. An example of 3 d.o.f. 3-D robot DDArm illustrates the results.     

Dynamical couplings analysis of rigid manipulators

A method for investigation of the dynamical couplings between the manipulator links is presented in this paper. The method is based on quasi velocities introduced originally by Hurtado (2004) and on a diagonalized Poincaré form of the Hamel representation described by Sovinsky et al. (2005). Some observations and a heuristic algorithm are proposed to determine the dynamical couplings between the rigid manipulator links. The presented strategy can be used in the design phase of manipulators and it can be realized via a simulation test. It is validated here on a 3 d.o.f. spatial manipulator.     

Adhesives to empower a manipulator inspired by the chameleon tongue

Existing grasping technologies have persistent challenges with unstructured objects and environments,highlighting an increasing demand for methods that conform to various application scenarios.Inspired by the chameleon tongue,a soft-contact grasping manipulator empowered by a class of adhesive gels has been demonstrated.The adhesives enable the manipulator to rapidly and strongly adhere to diverse substrates with varied surfaces,shapes and sizes,also to release objects under mild conditions.The robustness of such adhesive gels was highlighted with the remarkable recyclability,broad temperature tolerance and long-term stability.Furthermore,a general approach was developed to reconcile the contradiction of simultaneously enhancing their interfacial adhesion and cohesion strength that exists in conventional glues.We anticipate that this work will offer a strategy of developing adhesive materials and pave the way towards new applications of soft materials in the emerging fields of soft robotic devices and smart manufacturing.     

排爆机械臂逆运动学的模拟退火神经网络算法

采用传统方法计算没有封闭解的机械臂的逆运动学运算量大、精度无法保证,对于复杂结构很难满足实时精确控制的要求。六个并行三层双隐层前馈神经网络被设计用来解决排爆机器臂的逆运动学问题,神经网络的应用受到输出误差的限制,需要减小网络输出误差。针对机械臂结构,以神经网络输出为初始值,对网络输出关节变量进行实值编码,采用分离位姿模拟退火算法对的机械臂末端位置、姿态分别进行优化。仿真结果显示,该方法有效的减小了网络输出误差,在运算结果精确性和运算速度方面满足排爆机械臂求逆运动学解的要求。     

基于几何分析的机械臂运动路径规划问题研究

设计了分步-搜索逼近算法,解决了六自由度机械臂的位姿逆运算问题,使机械臂完成指尖到达指定目标点、沿指定曲线轨迹移动等动作.针对避障问题,设计了扫描截面的避碰算法.在此基础上,给出了机械臂作业的通用算法,并对所建模型和算法进行了实际检验.最后,分析了机械臂的设计参数,并提出了改进建议.