Exponential parameter and tracking error convergence guarantees for adaptive controllers without persistency of excitation
A modified concurrent learning model reference adaptive control method is proposed to guarantee the global convergence of parameter estimation error without assuming perfect knowledge on the state derivative. The value of the basis function vector is stored and manipulated in the original concurrent learning model reference adaptive control scheme. But in the proposed method, the time integral of the basis function vector is used instead of the basis itself to construct the history-based adaptation signal. By this modification, a smoother or an observer to get the state derivative estimate is not required in the proposed method unlike the original scheme. Therefore, the implementation of the proposed method is simpler than the previous one.