零超越量之積分器初始值補償方法與應用

No Thumbnail Available

Date

2009-08-01

Authors

呂有勝

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

本計畫提出一種積分器初始值補償方法,使系統響應無超越量,並將此設計方式運用於實際的物理系統。積分控制常用於降低穩態誤差,但同時導致暫態響應之超越量。以往之零超越量響應研究假設受控體之初始狀態為零;然而在模式切換控制系統中,當某個控制器被啟動時,受控體之初始狀態通常不為零,先前之零超越量研究結果不適用此種狀況。模式切換控制系統之相關研究也未能保證系統響應無超越量。本計畫提出之補償方法適用於受控體之初始狀態不為零之情況,能保證系統響應無超越量。 第一年計畫針對PID控制器之積分器初始值補償,使系統響應零超越量。先前模式切換控制之初始值補償研究限定控制器之初始狀態與受控體之初始狀態需為線性關係;由本計畫之推導發現,應補償之積分器初始值與受控體之初始狀態為非線性關係,這是與先前之研究不同之處,也說明先前之研究無法保證系統響應無超越量。第一年計畫將提出之補償方法應用於光碟機之聚焦伺服跳層控制(Layer-Jump Control)與單軌跳躍控制(One-Track-Jump Control)。 對參數不確定性大且外界干擾嚴重之受控體而言,第一年研究之PID控制器可能顯得強健性不足,因此在第二年計畫中導入滑動模式控制(Sliding- Mode Control, SMC)。更確切而言,第二年計畫將修改傳統的積分型可變結構控制(Integral Variable Structure Control, IVSC)設計,藉由提出之積分器初始值補償,不但使得滑動模式存在於整個系統響應過程中,增加系統強健性,而且使系統響應零超越量。第二年計畫擬將提出之補償方法應用於線性運動平台,以不同大小之負載與不同之負載速度,測試與比較傳統SMC、傳統IVSC、與本計畫提出之控制方法之性能。 本計畫以實驗方式,應用於實際物理系統,以證實提出之設計方法之實用性與有效性。透過這些應用實驗,本計畫訓練學生建構DSP/FPGA (Digital Signal Processor / Field Programmable Gate Array)為基礎之伺服控制系統,期能培育技術人才,並促進產學交流。
This proposal presents controller designs for nonovershooting responses by compensating for the initial value of integral feedback. Previous studies on nonover- shooting responses assume that the plants are initially at rest. However, in most “mode-switching control systems” where various controllers are activated interchangeably, the plants have nonzero initial conditions when some controller is just activated. For “mode-switching control systems”, the results on nonovershooting responses have not been found so far. Distinct from the previous researches, the method presented in this proposal ensures nonover- shooting responses even when the plants are not initially at rest. The first-year proposal deals with the initial value compensation for the integral feedback of a PID controller, in which the integral feedback is initially set to a value so that the subsequent response has zero overshoot. In the proposed scheme, referred to as Zero-Overshooting PID (ZOPID), the initial value that should be assigned to the integral feedback is found to be a nonlinear function of the plant’s initial state. This result is different from those obtained in the previous initial-value-compensation schemes that constrain the controller’s initial state to be a linear function of the plant’s initial state. For validations, the proposed scheme is to be applied to the Layer-Jump Control and the One-Track-Jump Control of an optical disk drive. The second-year proposal introduces the sliding-mode technique in order to improve the robustness of the proposed compensation scheme. More specifically, the standard Integral Variable Structure Control (IVSC) is redesigned with double integral feedback. One integrator’s initial value is assigned to have a sliding mode exist throughout an entire response, enhancing performance robustness. The other integrator’s initial state is assigned so as to guarantee nonovershooting responses. This proposed scheme is referred to as Zero-Overshooting IVSC (ZOIVSC). The ZOIVSC is to be applied to a position control of a linear motion stage. Since all the schemes are practically applied to physical mechatronic systems, the graduate students participating in this project can receive advanced training in Mechatronics and System Integration.

Description

Keywords

Citation

Collections