【摘要】：切換系統(tǒng)作為一類特殊的混合動(dòng)態(tài)系統(tǒng),被廣泛應(yīng)用于生物工程、航空航天工程和汽車工業(yè)等領(lǐng)域.在對(duì)切換系統(tǒng)控制問題進(jìn)行研究時(shí),大都假設(shè)模型切換和控制器切換的時(shí)刻是一致的,即同步切換.然而在實(shí)際應(yīng)用中識(shí)別運(yùn)行的子系統(tǒng)和請(qǐng)求相一致的控制器可能需要一些時(shí)間,控制器切換通常也要滯后相應(yīng)子系統(tǒng)切換,從而導(dǎo)致異步切換.針對(duì)如切換系統(tǒng)這類受限復(fù)雜系統(tǒng),如何在獲得控制器設(shè)計(jì)的系統(tǒng)性方法的同時(shí)實(shí)現(xiàn)性能指標(biāo)最優(yōu),控制算法可行和閉環(huán)系統(tǒng)穩(wěn)定是現(xiàn)代控制理論的一個(gè)極具挑戰(zhàn)性的問題.模型預(yù)測(cè)控制(Model predictive control,MPC)能夠有效處理系統(tǒng)約束和參數(shù)不確定性等因素,成為應(yīng)用較為廣泛的現(xiàn)代控制方法,也是解決上述問題有效的控制策略.本文研究了若干切換系統(tǒng)的模型預(yù)測(cè)控制問題,主要內(nèi)容概述如下:1.針對(duì)一類干擾范數(shù)有界的連續(xù)時(shí)間切換廣義系統(tǒng),研究了帶有混合H_2/H_∞性能指標(biāo)的預(yù)測(cè)控制問題.在設(shè)計(jì)切換規(guī)則下,采用多Lyapunov泛函方法得到基于線性矩陣不等式表示的H_2/H_∞預(yù)測(cè)控制控制器存在的充分條件.保證采用H_2范數(shù)度量的閉環(huán)系統(tǒng)性能有較好水平同時(shí)具有H_∞干擾抑制水平λ.2.針對(duì)一類離散時(shí)間下的切換系統(tǒng),提出一種基于異步切換的預(yù)測(cè)控制方法.采用平均駐留時(shí)間(ADT)方法,在允許切入的子系統(tǒng)Lyapunov-like函數(shù)遞增的情況下,將擴(kuò)展的穩(wěn)定性結(jié)果運(yùn)用到預(yù)測(cè)控制中.選取采樣時(shí)間等于異步切換滯后時(shí)間以及最大Lyapunov-like函數(shù)遞增時(shí)間,在給定最大異步切換滯后時(shí)間下得到依據(jù)最小ADT的受限切換律,并通過求解quasi-min-max優(yōu)化問題得到控制器保證閉環(huán)系統(tǒng)是全局一致指數(shù)穩(wěn)定(GUES)的.
[Abstract]:As a special hybrid dynamic system, switched systems are widely used in the fields of bioengineering, aerospace engineering and automobile industry. When studying the control problem of switched systems, it is assumed that the time of model switching and controller switching is the same, that is, synchronous handoff. However, in practical applications, it may take some time to identify the running subsystem and the controller that is consistent with the request, and the controller switch usually lags behind the corresponding subsystem switch, which leads to asynchronous handoff. For constrained complex systems such as switched systems, it is a challenging problem in modern control theory that how to obtain the systematic method of controller design and achieve the optimal performance index at the same time, the feasibility of control algorithm and the stability of closed-loop system. Model Predictive Control (Model predictive control,MPC) can effectively deal with system constraints and parameter uncertainties. It has become a widely used modern control method and an effective control strategy to solve the above problems. In this paper, some model predictive control problems for switched systems are studied. The main contents are summarized as follows: 1. For a class of continuous-time switched singular systems with bounded disturbance norm, the predictive control problem with mixed H _ 2 / H _ 鈭，

本文編號(hào)：2322001