CASIA OpenIR  > 毕业生  > 硕士学位论文
时滞系统的辨识算法和模型预测控制
Alternative TitleIdentification Algorithm and Model Predictive Control for Time-delay Systems
王应应
Subtype工学硕士
Thesis Advisor林红权
2007-05-30
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Keyword时滞 辨识 模型预测 工程化 过程控制 Time-delay System Identification Model Predictive Control Engineerization Process Control
Abstract工业过程本身的物理特性存在延时,信号转换需要延时,网络通讯延时和计算延时等因素也无法忽略,这些都决定了时滞效应广泛存在,而普通的控制方法对时滞系统的控制效果并不理想。为了设计好的控制器,并应用到工程中,本文考虑了以下四个方面: 1) 分析了时滞系统的数学描述。时滞过程的数学描述不同于非时滞系统,在第二章,重点考察了时滞过程的数学描述方法,这是文章后面辨识和控制的基础。 2) 第三章讨论了从时滞系统的过程实验数据到时滞模型的辨识方法。从时域和频域两个方面,给出了时滞常数和时滞系统的一些辨识算法; 3) 可用于时滞系统的控制器很多,第四章给出了一个概述。然后,重点推导了预测控制算法,并分析了如何把预测控制算法应用到PID参数的整定上面,同时还给出了预测控制算法的一种便于实现的形式。 4) 为了将控制算法应用于实际工程,本文第五章涉及了控制理论工程化实现的一些内容。 所有算法均给出了Matlab的仿真实例,一些算法已在实验室平台上做过测试。本文观点和算法上的新颖之处包括: 1) 指出了时滞辨识和模型阶次辨识的相似之处;并将已有的一些非时滞系统的辨识算法推广到了时滞系统。 2) 推导了预测控制器的便于实现的形式;考察了PID参数的整定和预测控制算法之间的联系。
Other AbstractIn the industrial process control field, due to the aftereffect phenomena occurred in the real processes’ inner dynamics, the variety of transport delays, communication delays involved by the sensors and transducers, and the computational time, time-delay effects are inevitable. Nowadays, the performance of traditional control-lers is not satisfactory. Therefore, the design and application of the advanced control-ler which takes the time-delay into account are necessary. This thesis focuses on the four aspects: 1) The mathematical description of time-delay systems. Time-delay systems are quite different from non-delay systems. Chapter two gives the main descrip-tion of time-delay systems which is the base of the system identification and control. 2) Chapter three discusses how to identify the model from process data. Several estimation methods of delay time and the corresponding system are given from the two different regions: time region and frequency region. 3) In chapter four, a simple survey is provided to show the development of anti-delay controllers. Then, model predictive algorithm is deduced and a convenient realization structure is obtained. In addition, the mathematical relation between the PID parameters and model predictive algorithm is pre-sent. 4) For the sake of real world applications, the engineerization of control theo-ries is considered. At the end of each chapter, Matlab simulations are presented. A few algorithms have been tested on the control platform in our laboratory. Some novel viewpoints in this thesis are: 1) Point out the resemblance between model order estimation and delay time estimation. Extend the existing non-delay identification methods to deal with time-delay systems. 2) Deduce the realization-convenient structure of the model predictive algo-rithm. Study the relations between PID parameters tuning and model predic-tive algorithm.
shelfnumXWLW1062
Other Identifier200428014628015
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/7398
Collection毕业生_硕士学位论文
Recommended Citation
GB/T 7714
王应应. 时滞系统的辨识算法和模型预测控制[D]. 中国科学院自动化研究所. 中国科学院研究生院,2007.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20042801462801(1848KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[王应应]'s Articles
Baidu academic
Similar articles in Baidu academic
[王应应]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[王应应]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

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