Parallel extended state observer based control for uncertain nonlinear systems

Xilian Zhang,Tao Xu,Zhao Zhang,Zhisheng Duan
DOI: https://doi.org/10.1016/j.neucom.2023.126687
IF: 6
2023-08-18
Neurocomputing
Abstract:In this paper, a parallel extended state observer (PESO) is proposed by merging a series of different order linear tracking differentiators and first-order nonlinear differentiators. Both measurement noises and system uncertainties are taken into account in design procedures. A distinct feature of the parallel structure is that the observed information of each part is independent of each other, which effectively avoids the accumulation of the observation error of each part. The PESO is utilized to estimate the total disturbance, which reflects the combined impacts of internal uncertainties and external disturbances. Based on the observation results generated by the PESO, a composite controller is developed for output feedback control of uncertain nonlinear systems. The convergence performance of the proposed PESO and PESO-based controller is rigorously verified. Finally, some numerical simulation results are provided to verify the effectiveness and superiority of the proposed method.
computer science, artificial intelligence
What problem does this paper attempt to address?