南開科技大學圖書館 |

Language: 繁體中文

說明(常見問題)

南開科技大學

圖書館首頁

編目中圖書申請

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Model-based analysis and control for...

Parker, Robert Stephen.

Model-based analysis and control for biosystems.

紀錄類型:	書目-電子資源 : 單行本
正題名/作者:	Model-based analysis and control for biosystems./
作者:	Parker, Robert Stephen.
面頁冊數:	217 p.
附註:	Source: Dissertation Abstracts International, Volume: 60-08, Section: B, page: 4095.
Contained By:	Dissertation Abstracts International60-08B.
標題:	Engineering, Biomedical. -
電子資源:	Download PDF (下載PDF全文)
ISBN:	9780599425613

Model-based analysis and control for biosystems.
Parker, Robert Stephen.

Model-based analysis and control for biosystems. - 217 p.

Source: Dissertation Abstracts International, Volume: 60-08, Section: B, page: 4095.

Thesis (Ph.D.)--University of Delaware, 1999.

Biosystems offer a challenging set of modeling and regulation problems to the biological, medical, and engineering communities. The interests of the medical device and chemical process industries are turning toward biosystems for drug delivery, pharmaceutical, and commodity chemical applications. These inherently nonlinear and constrained systems require regulation. This work was concerned with the analysis of biosystems and the development of model-based methodologies for controlling these processes. The focus was on two case studies: insulin delivery to diabetic patients for regulating glucose concentration and cell biomass exit concentration control of a continuous-flow bioreactor.

ISBN: 9780599425613Subjects--Topical Terms:

1000005515
Engineering, Biomedical.

Model-based analysis and control for biosystems.
LDR:03626nmm 2200301 4500 001 1000004850
005 20070601084657.5
008 070601s1999 eng d
020 $a 9780599425613
035 $a (UnM)AAI9941033
035 $a AAI9941033
040 $a UnM $c UnM{me_controlnum}
100 1 $a Parker, Robert Stephen. $3 1000005968
245 1 0 $a Model-based analysis and control for biosystems.
300 $a 217 p.
500 $a Source: Dissertation Abstracts International, Volume: 60-08, Section: B, page: 4095.
500 $a Professor in charge: Francis J. Doyle, III.
502 $a Thesis (Ph.D.)--University of Delaware, 1999.
520 $a Biosystems offer a challenging set of modeling and regulation problems to the biological, medical, and engineering communities. The interests of the medical device and chemical process industries are turning toward biosystems for drug delivery, pharmaceutical, and commodity chemical applications. These inherently nonlinear and constrained systems require regulation. This work was concerned with the analysis of biosystems and the development of model-based methodologies for controlling these processes. The focus was on two case studies: insulin delivery to diabetic patients for regulating glucose concentration and cell biomass exit concentration control of a continuous-flow bioreactor.
520 $a A first-principles nonlinear model for human glucose-insulin interaction was developed using compartmental modeling. A linear Volterra series model was identified from input-output data. Model-based predictive controllers designed from this model and the Jacobian linearization of the nonlinear model were compared. Hard input rate and magnitude constraints were enforced, and the output constraint was handled in a "soft" formulation. Performance for meal disturbance rejection was superior to literature feedback-only control algorithms. Inter- and intra-patient uncertainty was an important issue, and this variability was included in the algorithm through an adaptation mechanism. Up to 24% reductions in meal disturbance rejection tracking error were demonstrated versus the fixed parameter controller. Continuous and discrete-time robust Hinfinity controllers were synthesized, and significant patient variability could be tolerated in disturbance rejection scenarios.
520 $a A nonlinear Volterra model was identified from a literature continuous-flow bioreactor model using tailored plant-friendly input sequences to minimize data requirements, and this model was projected onto a Laguerre basis for model order reduction and coefficient smoothing. A nonlinear internal model controller was synthesized via model partitioning. The controller and closed-loop are demonstrated nominally stable across the optimal point for feasible references using the structured singular value and conic sector bounding of nonlinearities. Input magnitude constraints limit the performance of the direct synthesis design. An analytic solution to the nonlinear model predictive control squared 2-norm objective function was derived, including dynamic error penalization and manipulated variable weighting. This controller tracks the optimal constrained solution for all setpoints. Recursive least-squares was employed to update the model coefficients on-line as process behavior changes.
590 $a School code: 0060.
650 4 $a Engineering, Biomedical. $3 1000005515
650 4 $a Engineering, Chemical. $3 1000005684
690 $a 0541
690 $a 0542
710 2 0 $a University of Delaware. $3 1000005603
773 0 $t Dissertation Abstracts International $g 60-08B.
790 1 0 $a Doyle, Francis J., III, $e advisor
790 $a 0060
791 $a Ph.D.
792 $a 1999
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9941033 $z Download PDF (下載PDF全文)