LAUSR

We propose a gain-adaptive robust controller for position tracking of a brushless direct-current electric motor system. The controller has two design loops: an inner highly robust loop and an outer gain-adaptive loop. The inner loop is based on a backstepping framework to eliminate certainties in the system dynamics, while problem of internal and external disturbances is dealt with using a robust term directly synthesized from the tracking error. All control gains of the control signal produced by the robust loop are tuned online by a new gain-learning mechanism for the desired control performance. The gain dynamics are activated by special combinations of state control errors and deactivated by saturation functions. The stability of the closed-loop system is strictly proven by Lyapunov-based approaches. To verify the performance of the controller in a real-time system, a two degree-of-freedom robot leg was fabricated with the investigated motor system and experiments were carried out. The results obtained confirm the robustness, adaptation, fast response, and high accuracy of the proposed method.