摘要
PID控制和基于模型的先进控制广泛应用于工业过程中,为了很好的设计PID控制器或基于模型的先进控制器,确保取得较好的控制效果,需要准确的获取被控过程的临界信息及模型参数。本文采用继电反馈方法,对可以近似为一阶惯性纯滞后环节的工业对象施加继电激励信号,根据对象的振荡输出曲线,进行模型参数和临界信息的辨识。依据数学推理和大量仿真实验观察,本文改进了模型参数的辨识公式。仿真数据表明,相比理想继电反馈,饱和继电反馈对临界信息的辨识精度更高;在临界信息精确的情况下,依据改进的模型参数辨识公式,饱和继电反馈能够很好的辨识出被控对象的模型参数,从而为PID控制器或基于模型的先进控制器的设计提供了一个定量的参考依据。
PID control and model-based advanced control strategy is widely used in industrial process. To design PID controller or model-based advanced controller and achieve good performance, accurate ultimate parameter and model parameter is needed. In this paper, relay feedback method is employed. Relay signal is used to the industrial plant which can be approximated as first order plus dead time model. Model parameters and ultimate parameter can be identified from the oscillating curve of the plant. The model parameter identification formula is improved under mathematical derivation and lots of simulatiun experiment. Compared with ideal relay feedback, simulation results show that ultimate information can be identified more accurate through saturation relay feedback. Under accurate ultimate information and modified model parameter identification formula, model parameter can be identified accurately through saturation relay feedback. The exactly identification parameters provide a quantitative reference to design PID controller or model-based advanced controller.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2012年第9期1107-1110,共4页
Computers and Applied Chemistry
基金
国家自然科学基金资助项目(61004028,61272064)
中央高校基本科研业务费项目(WH1014013)
上海市教育委员会科研创新项目
关键词
饱和继电反馈
模型辨识
参数整定
saturation relay feedback, model identification, parameter self-tuning
