Transient Performance Analysis of Adaptive Control with Control Architecture Modification

Adaptive control is a promising control technique that could handle uncertainties arising in system degradation, actuator failure or structural damage. However, two major issues, namely poor transient performance and lack of robustness, hinder the application of adaptive control in practice. First, a modified adaptive control algorithm with tracking error feedback is proposed to suppress high-frequency oscillations while improving the tracking performance at the initial phase of the closed-loop system and give a formal stability proof by Lyapunov direct method. Then, we show that the large adaptive rate can reduce the bounds of tracking errors and control signals, but it can also exacerbate the high-frequency oscillation of the control signal. Moreover, increasing the tracking error feedback gain can reduce the bound of the tracking error while increasing the damping ratio of the control signal, which is equivalent to smoothing the control signal. Finally, numerical simulations are resorted to demonstrate that the modified MRAC control method can achieve fast adaption with guaranteed transient performance.