Uncertainty and Disturbance Estimator

This project was part of my undergraduate research work. The project dealt with the design of a discrete-time design of an Uncertainty and Disturbance Estimator for robust control. Despite the ease of implementation of discrete-time strategies, almost all prior work on UDE is for designing continuous-time control laws, with no general, complete research for discrete-time design. To design an appropriate discrete-time control law, a novel digital filter similar to the original analog filter for disturbance estimation is designed, a discrete-time error-based control law is derived, and a detailed stability analysis is provided. However, most real-world, physical systems are nonlinear and continuous-time in nature. Thus, the techniques of sampling and digital analog (D/A) conversion are used, enabling the control of linear, time-invariant as well as a class of nonlinear, continuous-time systems using discrete-time UDE. The considered nonlinear system is for the phenomenon of wing-rock motion. Simulations results for the proposed techniques indicated highly accurate stabilization and tracking performance, with excellent disturbance rejection. In particular, it was seen that the proposed control law is less sensitive to initial values of the error when compared to the original continuous-time UDE law. This work was published in the International Journal of Robust and Nonlinear Control.

This project was further extended to design a robust observed for controlling robot manipulators. The design resulted in a complete closed-loop, robust, controller–observer structure. The observer incorporates the estimate of the overall uncertainty associated with the plant, in order to mimic its dynamics, and the control law is generated using an auxiliary error instead of state tracking error. A detailed qualitative and quantitative stability analysis was carried out, and simulations were performed on the two-link robot manipulator system. Further, a comparative study with well-known control strategies for robot manipulators was also completed. The results demonstrated the efficacy of the proposed technique, with better tracking performance and lower control energy compared to other strategies. This article is currently under review. The open-access article can be found here.