Extended-State-Observer-Based Adaptive Control of Electrohydraulic Servomechanisms Without Velocity Measurement

Abstract

Velocity signal is difficult to obtain in practical electrohydraulic servomechanisms. Even though it can be approximately derived via numerical differentiation on position measurement, the strong noise effect will greatly deteriorate the achievable control performance. Hence, how to design a high-performance tracking controller without velocity measurement is of practical significance. In this paper, a practical adaptive tracking controller without velocity measurement is proposed for electrohydraulic servomechanisms. To estimate the unmeasurable velocity signal, an extended state observer (ESO) that also provides an estimate of the mismatched disturbance is constructed. The ESO uses the unknown parameter estimates updated by a novel adaptive law, which only depends on the actual position and desired trajectory. Moreover, the matched parametric uncertainty is also handled by online parameter adaptation and the matched disturbance is suppressed via a robust control law. The proposed ESO-based adaptive controller theoretically achieves an excellent asymptotic tracking performance when time-invariant modeling uncertainties exist. In the presence of time-variant modeling uncertainties, guaranteed transient performance and prescribed final tracking accuracy can also be achieved. The proposed control strategy bridges the gap between the adaptive control and disturbance observer-based control without using the velocity signal and preserves the performance results of both control methods while overcoming their practical performance limitations. Comparative experiments are performed on an actual servovalve-controlled double-rod hydraulic actuator to verify the superiority of the proposed control strategy.