A model consisting of three partial non-linear differential equations for describing the humidity and temperature of a thin sheet of material and of the temperature of the air flow as a function of time and position in an infrared dryer was reduced to a model more adaptable to direct digital control. To begin with the original model was discretized along the length of the dryer and then linearized. Models of orders ranging from 12 to 72 were analyzed for controllability and observability. Using Moore's method, the models thus obtained were further reduced to a controllable state space model of order 5 that produced satisfactory results and for which stability was preserved. Simulated open loop responses of the reduced model to discrete perturbations in emitter electrical power, inlet humidity and velocity of the web, produced satisfactory dynamic and steady-state responses when compared to responses of the original model subjected to similar perturbations. The steady-state responses showed absolute errors for humidity and temperature of the web less than 1.3% [g water/100 g] and 2.2°C, respectively.
Un modele decrit par trois equations differentielles partielles non-lineaires pour l'humidite et la temperature d'une nappe mince ainsi que la temperature de l'air en fonction du temps et de la position dans un sechoir infrarouge a ete reduit a un modele mieux adaptable a des fins de commande numerique. Le modele original fut tout d'abord discretise suivant la longueur du sechoir puis linearise. Des modeles d'ordre 12 a 72 furent analyses sur les plans de la contrǒla-bilite et de l'observabilite. Les modeles d'etat furent reduits davantage par la methode de Moore afin d'obtenir un modele d'ordre 5 pouvant produire des resultats satisfaisants tout en conservant la stabilite du modele reduit. Des simulations de la reponse de ce modele en boucle ouverte a des echelons dans la puissance electrique alimentee aux emetteurs radiants, dans l'humidite de la nappe entrant au sechoir ainsi que dans la vitesse de defilement de la nappe, indiquerent des reponses satisfaisantes en regimes transitoire et permanent lorsque comparees aux reponses du modele original soumis aux memes perturbations. Ainsi, en regime permanent, les erreurs absolues pour l'humidite et la temperature de la nappe a la sortie du sechoir furent inferieures a 1.3% [g eau/100g] et 2.2°C respectivement.
[1]
Masanobu Hasatani,et al.
HYBRID DRYING OF GRANULAR MATERIALS BY COMBINED RADIATIVE AND CONVECTIVE HEATING
,
1988
.
[2]
L. Silverman,et al.
Model reduction via balanced state space representations
,
1982
.
[3]
John H. Seinfeld,et al.
Mathematical Modeling of Packed Bed Reactors: Numerical Solutions and Control Model Development
,
1987
.
[4]
Normand Thérien,et al.
Modeling and simulation of the drying of thin sheets in a continuous infrared dryer
,
1994
.
[5]
Jules Thibault,et al.
STUDY OF HEAT AND MASS TRANSFER DURING IR DRYING OF PAPER
,
1994
.
[6]
Dale E. Seborg,et al.
Model reduction and the design of reduced‐order control laws
,
1974
.
[7]
E. Davison,et al.
On "A method for simplifying linear dynamic systems"
,
1966
.
[8]
B. Anderson,et al.
Model reduction for control system design
,
1984
.
[9]
M. Safonov,et al.
A Schur method for balanced-truncation model reduction
,
1989
.
[10]
Atsuo Watanabe,et al.
Factors Influencing Constant Drying Rate of Wet Granular Bed Irradiated by Infrared Radiation
,
1991
.
[11]
B. Moore.
Principal component analysis in linear systems: Controllability, observability, and model reduction
,
1981
.
[12]
Normand Thérien,et al.
Modélisation et simulation du séchage en continu des couches minces par rayonnement infrarouge
,
1990
.
[13]
Statistical Analysis of a Continuous Infrared Dryer
,
1991
.
[14]
Hirokazu Nishitani,et al.
Dynamic simulation of large systems.
,
1991
.