"Nominal Model Optimization for Improvement of Stability Robustness for Disturbance Observer-Based Control Systems,"
- Year of publication
- K. KONG and M. Tomizuka
- International Journal of Control, Automation, and Systems
Disturbance observer-based control systems often encounter the stability problem due to modeling uncertainties. In such cases, the disturbance observer (DOB) may have to be re-designed by narrowing the bandwidth of the Q-filter to enhance stability robustness, but this approach to stability enhancement deteriorates the performance of DOB. In order to improve robust stability while maintaining the performance of DOB, this paper proposes a method that manipulates the nominal plant model in the DOB; the parameters of the discretized nominal model are optimized to improve robust stability in the discrete time domain. For the optimization of nominal model, it is assumed that the closed-loop poles of DOB are subjected to multiplicative uncertainties, and the maximum allowable magnitude of uncertainties is utilized as the measure of stability robustness. Then, the proposed method changes the location of closed-loop poles to maximize the robustness margin. This paper provides a case study that includes experimental results.