Extended disturbance observer for nonlinear system with time-varying disturbance gain

In control engineering, the disturbance is a major problem that deteriorates system robustness. To address this problem, a disturbance observer (DOB) has been proposed and developed. In particular, a general method of estimating high-order disturbance, expressed as a time series expansion, was studied (H-DOB). However, systems to which H-DOB are applicable are limited to those in which disturbance gain is constant. This constraint degrades the generality of H-DOB because applicable systems are limited. Therefore, by expanding to a system with a time-varying disturbance gain, we propose a more general H-DOB. The DOB proposed in this paper redesigns the dynamics of the observer based on the structure of the H-DOB. The proposed DOB is applicable even if the system is highly nonlinear. This paper verifies the convergence of the proposed DOB through proof. Using this proof, the same characteristic error dynamics can be obtained, so the same design method can be applied. The simulation for the verification of the proposed method performs disturbance estimation of the extension system. This simulation proves that, compared to H-DOB, the proposed DOB has good disturbance estimation performance.

Automated summary

In this paper, a more general disturbance observer (DOB) is proposed to address the problem of disturbance deterioration in control engineering. The proposed DOB has broader applicability than existing methods, as it can be applied to systems with time-varying disturbance gain and even highly nonlinear systems. Through verified simulations, the proposed method demonstrates good disturbance estimation performance.