Failure analysis and fault tolerance of parallel robot manipulators

This research work is focused on analyzing the effect of active-joint failure on the kinematic performance in parallel robot manipulators, and on developing methodologies to modify the manipulator designs to ensure fault-tolerant robotic operation. In this research, formulations based on the Jacobian matrix analysis are developed to determine the effect of active-joint failure on the instantaneous velocity and static force at given manipulator poses. These formulations are used to develop a number of optimum post-failure motion trajectories based on various optimization scenarios, which provide the opportunity to control the robot operation after failure by reducing the task requirements. In addition, methodologies are developed to modify the design of parallel manipulators based on incorporating redundant active joints in their designs to recover the ability to move and apply forces in the task space after encountering active-joint failure. Optimization techniques are developed to select the optimum locations and the directions of the redundant joints for maximized robot performance.; The contribution of this research is that it provides an effective methodology that is based on analyzing the change in the Jacobian matrix to investigate and simulate the effect of active-joint failure on the motion of the robot manipulator. It is shown that the analysis of null-space and the orthogonal subspace of the Jacobian matrix can effectively be used to formulate post-failure task trajectories and provide a criterion to determine whether or not the manipulator is capable of moving or applying force in a given direction in the task space. The analysis also provides formulations that can easily be used to generate optimum end-effector trajectories for performing partial tasks after failure. Having the ability to perform partial tasks after encountering failure is very important in applications in which the undesired abortion of the robotic operation can endanger life, cause dramatic loss of property; or lead to serious environmental damage.; In addition, this research provides designers with an efficient procedure to determine the optimum locations and directions of backup joints that can be incorporated in the design of parallel manipulators to provide them with the ability to retain their full operational capability after encountering active-joint failure. This is important in cases where the manipulator needs its full operational capability to prevent the catastrophic consequences of failure. A developed numerical procedure, which is based on vector algebra, can efficiently be used to determine the optimum backup-joint directions that maximize the manipulator post-failure performance at a large number of manipulator poses covering a large workspace.