Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/85839
DC FieldValueLanguage
dc.contributorDepartment of Mechanical Engineering-
dc.creatorWu, Wei-
dc.identifier.urihttps://theses.lib.polyu.edu.hk/handle/200/2930-
dc.language.isoEnglish-
dc.titleModel-independent control of a class of uncertain systems-
dc.typeThesis-
dcterms.abstractThis research aims to develop a new control approach for a class of uncertain systems. Despite intensive research efforts of many researchers, system uncertainty in model parameters is still an existing problem in controller design. It is necessary to carry out research for a new controller, which is able to keep a stable closed-loop system and optimize the performance while requiring plant information as little as possible. The new control scheme is inspired by the idea of a PID self-tuning controller in the literature, which is able to maintain stable closed-loop regulation for a linear time invariant (LTI) system subject to unanticipated jumps of plant parameters and external disturbances. The controller depends on an assumption that the plant is open-loop stable with a full-row rank for the DC-feedback gain. The plant parameters are not required in the controller design. In this work, an extended high-order self-tuning controller is developed based on the same technique. Two typical versions are studied rigorously. One is called multi-absorber tuning control, which extends PID control by adding multiple resonator-absorbers. The other is called tuning control with a pseudo observer, which uses filter states to approximate the plant states within a user-defined bandwidth. Besides the stability issue, this work also addresses online optimization of the control performance. When the plant parameters are unknown, adaptive optimal control is the only available option. In this work, the simultaneous perturbation method (SPM) is adopted as an additional method to optimize the controller when the plant parameters are unknown. The integration of the stabilization algorithm and optimization control without plant parameters is the main contribution of this work, since this integration forms an independent module that is suitable to collaborate with any other available adaptive methods. The proposed control scheme is tested in both simulation and experimental examples. As far as simulation is concerned, within the assumption, all testing plant models are selected arbitrarily for the purpose of testing the controllers without knowing plant parameters. As for experiment, the control problem of flow induced vibration is used to test the practical applicability of proposed control method. The improved results in each example show the validation of the new control scheme. When studying the control example of flow induced vibration, this work also makes contributions on the understanding of coupled dynamics between flow vortex shedding and structure motion. At the preliminary stage, traditional model-independent methods are tested, such as resonator and variable structure control. However, the performance of these schemes is not effective enough. Besides the above-mentioned new controller, another control scheme is found to be effective, which uses a phenomenological low-order vortex oscillating model as guidance. The key idea is to effectively increase the nonlinear damping term in the model. A new multi-frequency perturbation method is developed, which is unique among all present control laws in fluid dynamics. Experimental results under resonant wind condition have shown its best performance in reducing flow-induced vibration when compared with other controllers. The robustness under different wind speed conditions has also been studied. All the findings have not only shown the effectiveness of proposed new scheme, but also validated the phenomenological model used to describe the complex interaction between vortex shedding and structure motion.-
dcterms.accessRightsopen access-
dcterms.educationLevelM.Phil.-
dcterms.extentx, 92 leaves : ill. ; 30 cm.-
dcterms.issued2005-
dcterms.LCSHHong Kong Polytechnic University -- Dissertations.-
dcterms.LCSHAdaptive control systems.-
dcterms.LCSHControl theory -- Mathematical models.-
Appears in Collections:Thesis
Show simple item record

Page views

44
Last Week
1
Last month
Citations as of Mar 24, 2024

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.