A Novel Hybrid Safety-Control Strategy for a Manipulator

Abstract

As robots tend to work cooperatively with humans in shared workplaces, safety as regards robot-human interactions has caused a great deal of concern in the robot community, and control strategies have become a hot topic in robotics research. In order to guarantee the robot's safety and continuous motions, this paper proposes a novel safety-control strategy, which is strictly conservative and which consists of a pre-contact and post-contact safety strategy. We adopt an optimal motion trajectory-planning method, by use of which the jerk, acceleration and velocity of the robot's motion can be limited and a time-optimal motion can be obtained as a post-contact safety strategy for a position-controlled manipulator. The optimal motion trajectory planning not only reduces the impact forces during the collision period, but also maintains the efficiency of the manipulator and preserves continuous motions. Next, we describe a novel collision detection method as a pre-contact safety strategy to avoid collisions. The method proposed here can compute security warning region to handle the effect of robot motion on collision detection and detect collisions between non-convex polygon soups. Finally, the control strategy is implemented for a 7-DOF humanoid manipulator and the experimental results demonstrate the validity of this novel hybrid safety-control strategy.