采用Smith预估器模型的时滞系统自适应控制
Adaptive control for time-delay systems adopting Smith predictor models
摘要点击 92  全文点击 64  投稿时间:2020-05-07  修订日期:2020-09-23
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2020.00251
  2021,38(3):416-424
中文关键词  时滞系统  Smith预估器  自适应控制  李雅普诺夫设计方法
英文关键词  time-delay systems  Smith predictor  adaptive control  Lyapunov design approach
基金项目  三菱电机研究开发基金项目(MEC–01)资助.
作者单位E-mail
郑仰东 三菱电机(中国)有限公司 研究开发部 zhengyd09@163.com 
中文摘要
      在实际生产控制系统中时滞现象相当普遍, 而传统的PID控制方法对时滞系统的控制很难达到满意的效果 且易引起系统的不稳定. Smith预估器是一种广泛应用于时滞系统闭环控制的有效方法, 但存在预估器模型和被控 对象参数不匹配时系统控制特性的劣化甚至发生振荡发散的问题. 本文提出了一种基于李雅普诺夫稳定性理论, 用 于改善Smith预估器参数失配时控制特性的模型参考自适应控制方法, 阐述了其基本原理并推导了1阶系统的自适 应律, 然后通过数值仿真计算进行了验证. 理论分析推导和仿真计算结果表明, Smith预估器和自适应控制方法的结 合明显改善了对时滞系统的控制性能, 防止了系统的不稳定性, 因而在一定程度上解决了Smith预估器存在的上述 问题, 取得了较好的效果, 有望在实际的时滞系统和设备的过程控制中得到广泛的应用.
英文摘要
      Time delay phenomena generally exist in actual industry control systems, but it is difficult to achieve a satisfactory control effect by using of the traditional PID control method which also easily causes the instability of the system. The Smith predictor is an effective method widely applied in closed-loop controls for the time-delay systems, however, there are serious problems that the mismatching of the parameters between the Smith predictor model and the plant will give rise to bad control characteristics or even the system oscillation and divergency. In this paper, we propose a Lyapunov-stability-theory-based model reference adaptive control scheme for improvement of the control characteristics for the Smith predictor methods when the parameter mismatch occurs. We describe the basic principle and derive the adaptive control laws for first order plant models. Then the numerical simulations are carried out to verify this proposal. The theoretical analysis and simulation results indicate that the combination of Smith predictor methods and adaptive control evidently improves the control performance of time-delay systems and prevents the system instability, therefore partly solves the above problems of the Smith predictor and acquires rather good effect. We expect it has a wide range of applications to actual process controls for time-delay systems and equipment.