文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 工程科技 > 机械/仪表 > adaptive pid controllers with specifications on phase and amplitude margins
侵权投诉

adaptive pid controllers with specifications on phase and amplitude margins

PID自适应控制

Adaptive PID Controllers with Specifications on Phase and

Amplitude Margins

Ya-Gang Wang, Xiao-Ming Xu

Abstract—This paper presents a new technique of adaptive PID controller with specifications on phase and amplitude margins. On the basis of process input and output data of normal operation in the control loop, the important frequency responses of process are estimated with signal decomposition and Laplace analysis, and then a second-order plus dead-time transform function is identified as well. Finally PID control parameters are calculated to satisfy both phase and amplitude margins. The adaptive procedure of PID controller does not need any prior knowledge of the process and previous controller while the control loop is still in the normal operation. Simulation examples are given to show both effectiveness and feasibility of the adaptive PID controller.

Keywords-Adaptive control; frequency response; amplitude margin; phase margin.

I. INTRODUCTION

I

N process control, more than 95% of the control loops are of PID type [1]. The main reason is its relatively simple structure, which can be easily understood and implemented in practice, and that many sophisticated control strategies, such as model predictive control, are based on it, The tuning of PID controllers are usually done by the seasoned engineers and technicians, that is laborious and time-consuming. Furthermore, because the real processes are different in dynamics, and have time delay and nonlinearity, the tuning of PID sometimes becomes very difficult. In result, many regulators are in practice poorly tuned, which represent a great economic cost for industry, so it is reasonable that PID auto-tuner is needed. Astrom and Hagglund proposed PID auto-tuning method based on relay feedback [2], which has successfully applied to the auto-tuning of PID controllers. This technique automatically extracts the process frequency response around an important frequency which is usually sufficient to tune a PID controller crudely for a simple process, overcoming the disadvantage of manual tuning method. A number of modifications and extensions of relay identification have been suggested in the literature [3]-[9], and some well-know automation companies developed their

auto-tuning products in the past years. However, PID auto-tuning need some additional input test signal to obtain process model without exception, such as step test or relay feedback test, and then calculate PID parameters based on certain algorithm, so they are usually suitable for initial operation and implementation of control loops. If the control loops are already in operation, but their performance become poor due to some reasons, thus normal operations have to be switched to test mode to re-identify the process model for update of PID parameters. It is inconvenient and usually forbidden in practical industrial processes. Moreover, the industrial processes and operation environment often change with time, so that previously tightly tuned PID controllers become poorly tuned ones. So PID controllers need continuous tuning during the operation to meet requirements of efficient production. However, the auto-tuning technology could not meet such requirements, thus the practical adaptive PID control technology becomes appealing. Adaptive PID control is different from PID auto-tuning function, because the adaptive procedure is implemented during normal operation without any disturbance or additional test signal, overcoming the drawbacks of auto-tuner.

This paper presents a new technique of adaptive PID controller with specifications on phase and amplitude margins. On the basis of process input and output data of normal operation in the control loop, the important frequency responses of process are estimated, and then a second-order plus dead-time transform function is identified as well. Finally PID control parameters are calculated to satisfy both phase and amplitude margins. Simulation example is given to show both effectiveness and feasibility of the adaptive PID controller.

II. ONLINE IDENTIFICATION METHOD

A. Process frequency response

In industrial processes, the set-points of PID control loops are often changed by operators or upper optimal system to meet efficient production and process optimization. Afterwards, process input and output signals vary from one steady state to another one, show as Fig. 1. The variation between two steady states includes dynamics characteristics of interest in the frequency domain. However, these input and output signals usually cannot be dealt directly with Fourier analysis since they are not absolutely integrable.

Manuscript received January 11, 2012. This work was supported in part by the National Natural Science Foundation of China under Grant 61074087 and the Innovation Program of Shanghai Municipal Education Commission of China under Grant 12ZZ144.

Ya-Gang Wang is with University of Shanghai for Science and Technology, Shanghai 200093, China (corresponding author, phone: 86-21-55271368; fax: 86-21-55274148; e-mail: ygwang@usst.edu.cn).

Xiao-Ming Xu is with University of Shanghai for Science and Technology, Shanghai 200093, China.

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email