Performance guaranteed fault-tolerant control of robotic manipulators under actuator faults and motion constraints

Most existing robust adaptive control designs for robotic systems only address the issue of modelling uncertainties and external disturbances. In this paper, we present a control method exhibits several salient features, such as robustness against modelling uncertainties, adaptation to unknown system parameters, tolerance to actuation failures, and obedience to motion constraints. It also ensures prescribed performance bounded (PPB) for any initial conditions, removing the shortcoming of the current PPB method. The proposed control scheme consists of two units: supervisory control and prescribed performance bounded (PPB) control. The role of the former is to drive the tracking error from any initial condition within the domain of interest to a residual set, so that the PPB control can be activated to ensure pre-given performance specifications. A soft connector is introduced to maintain continuous control action during the switching. Theoretical analysis and simulation verification confirm the effectiveness of the proposed strategy.