Quantitative analysis of critical limitation in using extended state observer
Active disturbance rejection control (ADRC) has been successfully widely applied. The extended state observer (ESO) is a crucial component of ADRC to deal with uncertainties in many control systems. Although the nominal stability of ADRC was proved by adopting a sufficiently fast observer with a performance recovery principle, it is difficult to be employed in practice because severe oscillation or even instability might be triggered if the fast observer was used. This paper investigates the reason behind this phenomenon within the framework of input time-delay sensitivity for a typical first-order system, which can provide an insightful understanding of ADRC. The positive root of the polynomial which determines the maximal input time-delay maintaining closedloop stability is quantitatively analyzed and the relationship among the bandwidth of ESO, other control parameters and the allowable input time-delay is thoroughly studied. Finally, numerical examples are presented to validate the corresponding theoretical results.
KeywordsActive disturbance rejection control extended observer input time-delay robustness
Unable to display preview. Download preview PDF.
- H. Sira-Ramirez, J. Linares-Flores, C. Garcia-Rodriguez, and M. A. Conteras-Ordaz, “On the control of the permanent magnet synchronous motor: an active disturbance rejection control approach,” IEEE Trans. on Control System Technology, vol. 22, no. 5, pp. 2056–2063, September 2014. [click]CrossRefGoogle Scholar
- Z. Gao, “Scaling and bandwidth-parameterization based controller tuning,” Proc. of the American Control Conf., pp. 4989–4996, 2003.Google Scholar