repository
불확실성을 가지는 비선형 시스템에 적용을 위한 강인 외란 추정 기반 제어 설계 방법
- Alternative Title
- Design of Robust Disturbance Observer-Based Control Methods with Applications to Uncertain Nonlinear Systems
- Author(s)
- Eom, Myunghwan
- Advisor
- 좌동경
- Department
- 일반대학원 전자공학과
- Publisher
- The Graduate School, Ajou University
- Publication Year
- 2016-02
- Language
- eng
- Keyword
- 시스템 불확실성; 펜듀봇 시스템; 외란 추정 기반 제어; LuGre 마찰력 추정기; 강 인 외란 추정기; 원자력 연구로; 외란 추정 기반 궤환 선형화 제어; 퍼지 기반 제한 기법; 외란 추정 기반 규정 성능 제어
- Alternative Abstract
- Practical control systems are always a_x000B_ected by system uncertainty which includes the friction force, parameter uncertainty, unmodeled dynamics, measurement noise, external disturbance, etc. These system uncertainties can degrade the control performance. Although much research has been conducted on the control considering the system uncertainty in the nonlinear mechanical systems, the disturbance observer-based control methods have been studied much rather recently. Recent study shows that the disturbance observer-based control for uncertain nonlinear systems to improve the control performance is still one of the major open problems. With this in mind, the focus of this dissertation is to develop a robust disturbance observer-based control strategy for the uncertain nonlinear systems including pendubot and nuclear research reactor system to achieve the satisfactory control performance. First, the friction force in the pendubot system has been formally introduced by both LuGre and dynamic friction models, which can cause the limit-cycle phenomenon around the unstable equilibrium point. In addition, the dynamic friction model in the pendubot system is verfi_x000C_ed through the simulation and experimental results. Here, we decompose a pendubot system into actuated and unactuated subsystem to compensate for the overall uncertainties at the same time. Then, we design the LuGre friction observer and the robust disturbance observer which ensure exactly compensation of its uncertainties in the pendubot system. We provide a comparison study which shows that the proposed control method provides more robust and effectiveness than previous swing-up and balancing control method in the presence of the system uncertainties. Second, a robust disturbance observer-based feedback linearization control (FLC) method combining an PI-like control law for uncertain research reactor system with the unmeasurable thermal feedback eff_x000B_ect is proposed. In addition, the input as actual motor can be obtained by the inverse power control system model in the practical point of view. Particularly, a fuzzy-based limiting method has been developed to constrain the power change rate within the critical value for safety reason, which makes it easy to be applied to the heavy industrial reactor system. Accordingly, we show that the proposed control method can guarantee the asymptotic stability as well as the improved control performance even in the presence of system uncertainties, unlike the conventional reactor control method. Finally, a disturbance observer-based prescribed performance control (PPC) law to deal with both system uncertainty and control performance is proposed, where PPC method is utilized to improve its performance such as transient response, steady-sate error, maximum overshoot, and convergence rate. Thus, the proposed approach is more e_x000B_ffective and can obtain the more satisfactory control performance compared with previous FLC. Therefore, the control performance can be signifi_x000C_cantly improved by compensating for the system uncertainty. Each result is supported through rigorous Lyapunov-based stability proofs, numerical simulations, and/or experimental demonstrations.
- Fulltext
- Appears in Collections:
- Graduate School of Ajou University > Department of Electronic Engineering > 4. Theses(Ph.D)
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
-
- License
- STATISTICS
- Total Visit :3,722,245
- Total Download :1,818
- Today View :5,448
