Utilisation optimale de l’espace de travail des robots parallèles en affrontant certains types de singularités

Abstract

The popularity of parallel robot has been increasing significantly in industry. There arethousands of them around the world, most frequently used for pick and place. They are fasterthan serials robots because their actuators are generally fixed to their base reducing the weight of the moving part. However, the growth of the popularity of these robots on the market is slowed down because of their reduced workspace. Indeed, the privileged approach is to avoid all singularities which results in a considerable reduction of the workspace. The main objective of this project is to increase the workspace of a five-bar parallel robot using path planning. An algorithm for path planning has been developed to take advantage of type 1 singularities in order to avoid type 2 singularities and also to optimize the trajectory by minimizing the cycle time. A control algorithm has also been designed to maximize the dynamic response of the robot. Those algorithms were validated on a two degrees-offreedom prototype developed in the CoRo laboratory of the École de technologie supérieure. The path planning algorithm determines the best of the four possible configurations for the two arms of the robot, for a desired Cartesian position (x, y) of the end-effector. The optimization process of the cycle time considers several constraints: the maximum velocity, the maximum acceleration, the maximum jerk and also the maximum torque based on the dynamic model of the robot. The dynamic model is also used in the control algorithm to achieve the specified performance criterion while ensuring the accuracy of the robot. The results show that the workspace of the robot is significantly increased in comparison to a robot of the same size that does not change configurations. Moreover, changing configurations also improves the cycle time. The type 2 singularities are avoided which ensures the rigidity of the robot throughout the workspace and allows it to do fast operations of pick and place where the path between the points is optimized. This research project has developed a new approach demonstrating that it is possible to work with singularities in order to maximize the dynamic performance and the workspace of this type of robot.

Date	13 Aug 2012
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Pascal Bigras (Supervisor) & Ilian Bonev (Co-supervisor)