上海交通大学学报(自然版) ›› 2015, Vol. 49 ›› Issue (06): 743-750.

• 自动化技术、计算机技术 • 上一篇    下一篇

间歇过程复合迭代学习容错保性能控制器设计

王立敏1,2,周东华1,朱城杰2   

  1. (1.清华大学 自动化系, 北京 100084; 2.辽宁石油化工大学 理学院, 辽宁 抚顺 113001)
  • 收稿日期:2015-03-10 出版日期:2015-06-29 发布日期:2015-06-29
  • 基金资助:

    国家自然科学基金(61490701, 61210012, 61290324, 61104058), 中国博士后面上基金 (2014M560972), 辽宁省高等学校杰出青年学者成长计划 (LJQ2014039)项目资助

Composite Iterative Learning FaultTolerant Guaranteed Cost Controller Design for Batch Process

WANG Limin1,2,ZHOU Donghua1,ZHU Chengjie2   

  1. (1. Department of Automation, Tsinghua University, Beijing 100084, China; 2. College of Sciences, Liaoning Shihua University, Fushun 113001, Liaoning, China)
  • Received:2015-03-10 Online:2015-06-29 Published:2015-06-29

摘要:

摘要:  针对具有执行器故障和状态时滞的间歇过程,主要考虑输出反馈与迭代学习控制相结合,即复合迭代学习可靠保性能控制器设计问题,将具有时滞的故障系统转化为等价的2DFM 模型的时滞系统,在等价模型基础上,提出了基于2D 动态输出反馈的鲁棒迭代学习可靠保性能控制策略,建立了依赖于时滞、确保系统稳定且具有最优性能的充分条件,并提出具有拓展学习信息的鲁棒迭代学习容错保性能控制的优化设计算法,同时,考虑了时滞对系统稳定性及性能的影响,实现了当系统发生执行器失效故障且在故障允许范围内时闭环系统的平稳运行,并具有最优的控制性能.另外,以注射成型过程保压段的喷嘴压力设计控制律为例验证了所提方法的有效性.

关键词: 间歇过程, 复合迭代学习, 可靠保性能控制, 执行器故障, 时滞依赖

Abstract:

Abstract: Focusing on the batch processes with actuator fault and state delay, a compound iterative learning reliable guaranteed cost controller combining feedback control and iterative learning control was proposed. The fault system with time delay was transformed to the delay system with the equivalent 2DFM model. Then, based on the equivalent model, to guarantee the realization of the system, a robust iterative learning reliable guaranteed cost method based on 2D dynamic output feedback was proposed with developed learning information. Timedelaydependent sufficient conditions guaranteeing the system stability and best performance was constructed. Furthermore, an optimal design algorithm with expanded learning information of robust iterative faulttolerant guaranteed cost control was derived, while taking the influence of time delay on system stability and performance into consideration. The closedloop system could run smoothly within the fault permissible range when actuator failure occurred. The application on the nozzle pressure in packing phase of injecting modeling process proves the effectiveness of the proposed scheme.
Key words: batch process; iterative learning; faulttolerant guaranteed cost control; actuator fault; delay dependent

中图分类号: