Table of Contents Table of Contents
Previous Page  88 / 100 Next Page
Information
Show Menu
Previous Page 88 / 100 Next Page
Page Background

CONTROLLO

tecnica

Aprile 2017

Automazione e Strumentazione

88

Tuning of PID controllers is one of the key aspects in control system deployment.

Tuning rules provide an easy way of obtaining values for a PID controller

given information about the process is available. In this note a very simple and

straightforward tuning procedure, called uSORT, is presented. It is based on first and

second order systems and provides optimal integrated absolute error performance

provided some degree of robustness is guaranteed.

Ramon Vilanova

Simple robust PID tuning

It is an indisputable fact that the

proportion-

al-integral-derivative

,

PID

, controller has been,

and it still is, the workhorse of control at process

industries. Even if more advanced control solu-

tions are permeating into the control rooms, the

PID continues to be the preferred solution at field

level. Therefore, the reliability of those advanced,

plant-level solutions, depends on the appropri-

ate performance at field level. We could state it

depends on the PID level.

One of the cornerstones of activity in PID con-

trol is that of adjusting the PID gains: tuning the

PID controller. Being this a basic need in con-

trol practice, the goal of obtaining simple, pow-

erful and as generalist as possible tuning rules

has been one of the major drivers of research

for improvements on PID control

[4]

. In this

note, attention is focused on simple expressions

for the PID gains with the purpose to reinforce

the use of advanced tuning approaches, as they

can take quite simple expressions.

PID based control system

The purpose of a control system is to obtain a

desired response for a given system. We consider

the typical feedback control system represented

in

υ

Figure 1

.

where

P

is the process to be controlled,

C

is the

controller,

r

is the set-point signal,

u

the control

signal,

d

is the load disturbance signal,

y

is the

process output,

n

represents the measurement

noise and

e:=r - y

is the control error (the differ-

ence between the process output and its desired

or reference value). Being

u

the command signal

generated by the controller, the application of a

PID controller is, essentially, the result of weight-

ing three different actions. The most usual PID

implementation is the so called ISA or Standard

PID that takes the form (using the expression in

the Laplace s-domain):

L’AUTORE

R. Vilanova, Departament de

Telecomunicació i Enginyeria de

Sistemes, Universita Autonoma

de Barcelona, Spain

IN ADDITION TO PERFORMANCE, ROBUSTNESS IS ALSO IMPORTANT FOR PID TUNING

Taratura dei PID semplice e robusta

La taratura dei controllori PID è uno degli aspetti chiave nella messa in funzione dei sistemi di controllo. L’uso di formule di taratura sempli-

fica questo aspetto laddove si ha un modello del processo. In questo articolo viene presentata una procedura semplice e intuitive chiamata

uSORT, basata su un modello del processo del primo o secondo ordine e che fornisce una prestazione che minimizza l’integrale del valore

assoluto dell’errore garantendo allo stesso tempo una adeguata robustezza.

Figure 1 - PID based control system