This article focuses on time-optimal trajectory planning for robots with flexible links. Minimum time trajectories along specified paths as well as time-optimal point-to-point motions, which avoid vibration excitation due to elastic deflections, are determined. This is achieved by additionally constraining parts of the generalized forces and generalized force derivatives, resulting from the elastic potential. Therefore, the dynamical robot model is obtained using the Projection Equation.In a further step, a reduced model with the most essential degrees of freedom and sufficient accuracy is introduced, resulting in a flat system. Utilizing this, a trajectory control with an exact feedforward linearization in combination with a feedback part, consisting of a motor joint as well as a joint torque control, is realized. This nearly ideal control is used for moving on the time-optimal trajectories. The optimization is conducted with respect to velocity, jerk and motor torques as well as the newly introduced constraints, computable due to the flatness of the system. Experimental results demonstrate the improvement concerning vibration avoidance of the considered robot. Furthermore, a comparison between the occurring bending stress and the maximum permissible bending stress shows that mechanical damage is prevented with the use of the additional constraints.

In this technical brief, a novel hydraulic drive for large forces and power ratings at relatively high operating frequencies combining variable displacement control and hydraulic digital control is introduced. Basic analog motion control is achieved via variable displacement pumps driving a first cylinder stage. Digital control is realized by switching additional hydraulic cylinder stages on and off to support the analog stage if high forces are needed. The control strategy corresponds to this hydraulic concept. It consists of a feed forward control, a switching logic for the digital booster stages and a feed back proportional-integral (PI) control for stabilization. The validity of this concept and of the control strategy are shown by experiments on a highly downscaled test rig.

No abstract

Kinematically redundant serial robots have become industrially important due their increased workspace and their inherent capability of null space motion resulting in remarkable adaptiveness to specific tasks compared to conventional, non-redundant manipulators. Attempting to increase the cost-effectiveness of industrial processes, introducing minimum-time trajectories may yield economical advantages due to reduced motion cycle times. This contribution presents a method that uses joint space decomposition and analytic inverse kinematics as well as standard optimization techniques to obtain minimum-time B-spline joint trajectories along prescribed task space paths for kinematically redundant serial robots. It is shown that the present method was successfully applied to a planar manipulator.

No abstract