Active self-protection control strategy of damaged robot joints

Whether a robot can autonomously regulate its motion posture to achieve self-protection after joint damage—similar to human self-protection mechanisms—and prevent further deterioration of the damage is a fundamental guarantee for its reliable operation in unattended environments such as outer space. To address the self-protection requirements of robots, this paper proposes an active self-protection control strategy with the combiniaiton of “motion compensation” and “active load reduction”. Unlike traditional fault-tolerant control that only focuses on the end-effector trajectory and positioning accuracy of the robot, this study focuses on the health state of the robot itself, aiming to avoid the exacerbation of damage by rationally regulating the motion posture and output torque of the damaged joints. To achieve the above objectives, a health factor is first introduced to quantify the torque-bearing margin of the damaged joints. Then, a three-layer control architecture is proposed. The first layer actively distributes motion tasks to healthy joints through weighted pseudo-inverse and null-space reconstruction technologies, realizing “motion compensation” for the damaged joints. The second layer adopts linear time-varying model predictive control (LTV-MPC). By performing sequential linearization on the nonlinear dynamics, the health factor is mapped into hard dynamic constraints, achieving “active load reduction” for the damaged joints at the physical level. The third layer employs a closed-loop feedback mechanism based on dynamic feedforward to ensure the precise execution of optimized commands. Simulation results show that even when the load-bearing capacity of the joint is only 30% of the original due to damage, the proposed strategy not only ensures that the robot completes the predetermined trajectory but also strictly limits the torque of the damaged joint within the safe margin. This effectively prevents further deterioration of joint damage and guarantees the operational reliability of the robot under degraded service conditions.