Kinematic Fault Tolerant Control for Redundant Robot Based on Joint Velocities Redistribution

This paper presents a new approach of joint velocities redistribution for kinematic fault-tolerant control of redundant robot. The suggested approach unlike the locking of failed joints can distribute special velocities to some joints of robot before or after the trouble take place, so that the robot can avoid suffering from worse joint positions and orientations, and its fault-tolerant ability will be improved. If some joints can’t be driven, their joint velocities should be given 0. Thus, to lock some joints is only a special example of the redistribution approach under of trouble. On the other hand, this approach can also be extended to kinematic optimization, such as the avoidance of joint angle limits, the avoidance of singularities, the avoidance of obstacles etc. Simulation verifies that this approach is very useful for real-time control because it has stronger effect than optimization control method.