No abstract

This paper presents simulation and optimization of a 3-D three cable driven manipulator. For some specific paths in the workspace volume, motions of the robot before optimization are simulated and tension in the cables and stiffness of the cable robot for various points on the paths are determined. By selecting a function which is related to the stiffness matrix of the robot and considering it as the objective function of genetic algorithm, vertical positions of the connection points on the base platform as the optimization variables for each point on the paths can be determined in a way that the related function of the robot is optimized and a tension reduction in all the cables can be observed for most of the paths. Finally, the results of using different kinds of objective functions derived from the stiffness matrix are presented for three different criteria: workspace volume, kinematic performance indices and actuating energy of the robot. The results indicate which function has the largest workspace volume with tension reduction property in all the three cables, which has the highest values for the kinematic performance indices and which has the lowest actuating energy.

This contribution presents the experimental investigations on design and fabrication of the dorsal fin of the shark robot-fish mechanism which is controlled using NiTi Shape Memory Alloy (SMA) actuators by a Fuzzy controller. The dorsal fin helps the shark not only to turn around, but also it helps shark to roll. This fin contains two main parts, a rigid part; is connected to the shark’s body and another part is mounted on the latter one which can be rotated. The rotation angle of the second part is controlled by two parallel SMA wires. The subtlety of this installation is the tension of the right wire, when the left one is actuated. In addition, a fuzzy controller was determined by considering the characteristics of the SMA wires and designated mechanism which has hysteresis response characteristics against heating and cooling.

This paper describes a novel design of force sensors arrangement for measuring ground reaction force on a humanoid foot to find the zero momentum point (ZMP) which is usually used as a stability criterion for a walking biped robot. Precise locating of ZMP highly depends on accurate measurement of reaction force distribution on the foot. Design and fabrication of a sole for a full scale humanoid robot, “Surena” is of interest. A rubber layer is placed on robot sole parallel to force sensors to reduce the effect of impact while robot is walking. Consequently, the force measured by sensors is not total force exerted by the ground. Several foot interaction condition with ground have been modeled using finite element method and force distribution in sole rubber and sensors have been obtained. The resultant is a model which predicts the total force distribution based on four point force measurement. Several experimental situations have been tested and forces have been measured. Experimental results have shown that numerical model could predict force distribution in a good way and location of ZMP has been found successfully.