节点文献
输出随机分布控制系统的控制回路性能评估方法研究
Research on Performance Assessment Method of Control Loop of Output Stochastic Distribution Control System
【作者】 王曦;
【导师】 周靖林;
【作者基本信息】 北京化工大学 , 控制科学与工程, 2014, 硕士
【摘要】 控制器设计是控制回路的主题,然而控制回路是否按设定好的目标工作或控制器设计是否合理则是控制回路性能评估需要回答的问题。高斯系统和确定系统的控制回路性能评估已有许多满意的成果。本文的主要工作为研究针对输出随机分布控制系统的控制回路性能评估方法,分别从随机分布控制系统的跟踪稳态性能,随机性能展开介绍。并在输出随机分布控制系统领域的理论基础上,结合传统高斯系统的某些性能评估方法,对两者之间的共性和区别进行深入分析,整理出一些适合输出随机分布控制系统的控制回路性能评估方法,主要有:(1)对输出随机分布控制(SDC)系统的跟踪控制回路的稳态性能评估方法展开了分析。首先基于线性矩阵不等式和增广控制设计了一类输出PDF控制系统的形状跟踪器,在此基础上给出了一类输出PDF控制系统的形状跟踪回路的稳态跟踪精度的度量方法。(2)分析了基于Shannon熵的最小熵控制器的缺陷并定义了有理熵(RE)。当PDF采用B样条逼近时,给出了相应的均值约束下的RE最小值等相关结论并以此为基准给出了输出SDC的不确定性的性能评估方法。(3)最后对本课题研究的输出随机分布控制系统的性能评估方法的工作进行总结和展望。
【Abstract】 Controller design is the theme of the control loop. However, whether the control loop is running in a good condition which we expect, and the controller design is reasonable or not, is the main question which the performance assessment of the control loop needs to be answered. The control performance assessment has made a series of substantial achievements in Gaussian systems and deterministic systems.The main work of this paper is to study the performance assessment method of the output stochastic distribution control (SDC) system. It can be introduced from three aspects, including static performance and stochastic performance. Through the analysis with the similarities and differences in Gaussian systems and SDC systems, we finally build several innovative methods based on the developed fundamental theory in the field of output stochastic distribution system and the performance assessment methods developed in the general Gaussian system. The main contribution of this article can be summarized as follows: (1) Firstly, we consider the steady state performance of tracking control of the SDC system. We design a kind of tracking controller based on LMI and augmented control for the stochastic process. Then we propose a performance assessment characterize the steady-state tracking accuracy for the output PDF control system.(2) A new type of entropy named rational entropy (RE) was adopted for our research by analyze the defects of the minimum entropy controller based on Shannon. Then we can draw a theorem about the minimum RE with mean constraint analysis based on the PDF approximated by linear B-spline model. Then, we propose the performance assessment methods for characterizing the uncertainty performance of the SDC system.(3) Finally, a brief summary and outlook of performance assessment for the output SDC system was given.
【Key words】 Performance assessment; Output stochastic distributioncontrol; Steady state performance; random performance; Minimumentropy control;