Soft tracking control for input constrained systems

Actuator saturation, a common nonlinear problem in practical control systems, can not only degrade the performance but may also lead to system instability. Due to the perturbation of large disturbances or inappropriate trajectory planning, the system state may deviate farther away from its reference than the controller can react to properly without becoming saturated. A new "soft tracking" technique proposed in the paper provides a system with an auxiliary feasible reference such that the controller can be recovered from the saturation and the output can gradually converge to the original tracking profile. In contrast to regular tracking control designs, the auxiliary reference used in calculating the control law is online generated by incorporating the current system states and the original desired trajectory in the design.