FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

PROCESS INSTRUMENTATION AND CONTROL

Moreira, Rosana G.
Department of Agricultural Engineering, Texas A&M University, USA

Keywords: Feedback, PID, SISO, transmitter, digital control, PLC, sensors,

disturbances, loads, software

Contents

1. Introduction
2. Background
2.1. Variables in a Control System
3. The Control Problem

S
TE S
3.1. Feedback Controller
3.2. Open-loop and Feedforward Control Systems

R
AP LS
4. Instrumentation
4.1. Measurement Transducer
4.2. Measurement Devices or Sensors
C EO
4.3. Final Control Elements and Regulators
4.4. Transmission Lines
5. Computer-Controlled Systems
5.1. Digital Control Hardware
E –

5.2. Programming Language

H

5.3. Programmable Logic Controllers

PL O

6. Future Trends
7. Future Directions
M SC

Glossary
Bibliography
Biographical Sketch
SA NE

Summary

This article presents an overview of the area of instrumentation and control as applied
U

to food process systems. It is divided into three major areas: control, instrumentation,
and digital computer control.

Automatic control in food process systems is necessary to maintain final product

quality, improve process efficiency, and reduce waste. Limited understanding of the
physicochemical interactions occurring during processing, as well as large and
unpredictable variations in raw material composition and characteristics, in addition to
large dead time, make it difficult to design proper controls for food process systems.

The combination of empirical modeling together with fundamental modeling will

significantly improve the understanding of food process systems dynamics. With
improved, faster computers and the development of new on-line smart sensors,
advanced control techniques will be easily implemented to handle most of the
complexity of food process systems. Great attention should be given to the

©Encyclopedia of Life Support Systems (EOLSS)

FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

advancement of control techniques, such as closed-loop performance monitoring,

statistical process control, fuzzy logic, and neural network controls.

The progress in sensors and measuring devices has improved significantly in the last
decade. Today, smart-sensors can measure, self-compensate for disturbances, and
even amplify signals. Advances in machine-vision technology will contribute to better
characterizations of product properties on-line.

Another important area of ongoing improvement is in digital computer control. The

integration of Places and Doss, or LC hybrids will efficiently handle the complexity
of combining the batch and continuous controls required in the food industry. New
programming tools are also being developed to improve control applications and
communications.

1. Introduction

S
TE S
R
AP LS
The automatic control of food process systems helps improve final product quality,
increases process efficiency, and reduces waste from raw materials. Food processes are
generally multiple input/output systems that involve complex interactions between
C EO
process inputs and outputs (see Food Process Engineering). They are generally
characterized by strong relationships among mass, energy, and momentum transfer, in
addition to complex physicochemical transformations that include gelatinization of
starch, denaturization of proteins, and browning reactions. Such changes are influenced
by the chemical composition and physical state of the raw materials used and by the
E –

process conditions.
H
PL O

It is difficult to design proper controls for food process systems because of the limited
understanding of physical and chemical interactions occurring during food processing.
M SC

In the past, lack of on-line sensors has prevented direct control of product quality
parameters. Instead, previously proposed control schemes controlled product quality
indirectly through secondary process variables.
SA NE

Raw materials used in food processes are mainly of biological origin and their
compositional and physical nature can vary considerably (see Engineering Properties of
U

Foods). Such variations introduce significant, unmeasurable disturbances to the process

that can make manual control unreliable. In addition to the complexity caused by raw
material variability, food process systems also exhibit larger dead time and are typically
non-minimum phase systems. For systems with significant time delays, improved
control performance over PID (proportional-integral-derivative) controllers has been
achieved with the Smith predictor and Minimum Variance controller schemes.

A new class of controllers, Model Predictive Controllers (MPC) or more specifically,

Generalized Predictive Controllers (GPC), has been found to overcome the limitations
associated with Minimum Variance and the Smith predictor control schemes. The
Generalized Predictive Controller with an ARX (auto-regressive with exogenous input)
model has successfully been employed to control a twin-screw food extruder and a
continuous fryer (see Extrusion, and Frying). Dynamic Matrix Control (DMC) has been
applied to a continuous frying process achieving good results.

©Encyclopedia of Life Support Systems (EOLSS)

FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

A combination of GPC with an on-line identification routine has been shown to be

capable of providing stable control to processes with variable parameters and variable
dead times, and with model orders that change during the process. This type of
controller is referred to as adaptive. In food processing systems, the operating
environment can drift to the extent that raw material properties (particle size, moisture
content, composition, etc.) will vary and machines wear or deteriorate due to long-term
use. Constantly changing dynamics, variable dead times, and model parameters that
change depending on the direction and size of control movement are typical in many
food processes. These typical disturbances are not easily modeled because on-line
sensors are not readily available to measure them. Therefore, fixed controllers may fall
short in compensating for such disturbances over time, making adaptive controllers
more attractive for food process systems. Adaptive control has been successfully
applied to continuous flow grain dryers and food extruders.

One of the difficulties in applying adaptive control to food process systems is that in

S
TE S
addition to normal wear of machine parts, abrupt failures caused by actuator and sensor

R
AP LS
malfunctions and failures arising from the process itself are common. When these faults
occur, they must be detected rapidly and be distinguished from wear, so that the process
can continue to run safely. In conventional adaptive control, the process model
C EO
parameters are recursively estimated and used to recompute the controller parameters in
every sampling instance.

However, estimator variations due to disturbances and parameter tracking are passed to
the controller. Closed-loop performance monitoring is a technique that can be used to
E –

update controllers only when the system has changed and the parameters converged
H

with sufficient confidence. This control technique would be of great help for food
PL O

processing systems and needs to be investigated further.

M SC

2. Background
SA NE
U

Figure 1. Manipulated, controlled, and disturbance variables in a food extrusion process.

©Encyclopedia of Life Support Systems (EOLSS)

FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

Consider, for example, a schematic of a food extrusion process with all the important
variables needed for process control shown in Figure 1. From the production point of
view, a process can be defined as a progressively continuing operation that consists of
controlled actions systematically directed toward a desired product. Food extrusion is a
process that consists of feeding, mixing, cooking, forming, and puffing.

The major functional ingredient in expanded products is starch. During extrusion, starch
is plasticized with water and subjected to a specific mechanical and thermal energy
treatment as it moves through the extruder. The physicochemical changes that occur in
starch at high shear, low moisture, and high temperature conditions dictate final product
quality.

Expansion at the exit of the die is driven by internal steam pressure in the extrudate. At
the die inlet, the process pressure and temperature maintain the vapor in a superheated
condition. When the product reaches the die exit, the sudden release of pressure causes a

S
TE S
flushing of superheated vapor into steam. The large pressure differential between the

R
inside of the extrudate and the surface causes steam to rapidly expand and migrate

AP LS
toward the surface of the product. The steam exits into the atmosphere through the
product porous spaces created during the expansion process.
C EO
2.1. Variables in a Control System

From the control point of view, the number of manipulated variables must be at least as
large as the number of controlled variables. The controlled variables should measure
E –

product quality directly or strongly affect it. Manipulated variables should have a large
H

effect on the controlled variables (large gain), rapidly affecting the controlled variables
PL O

(minimum delay, small time constant) directly rather than indirectly.

M SC

Extrusion process parameters include raw materials, feed rate, water rate, screw speed,
barrel temperature profile, and the screw/die configuration. They are considered
independent operating variables, with the exception of the screw/die configuration,
SA NE

which is typically fixed for a specific application. Changes in these operating

parameters will cause changes in the dependent process variables, such as die product
temperature, die pressure, and viscosity, as well as product quality attributes (PQA),
through changes in the specific mechanical (SME) and thermal energy input during the
U

products residence in the extruder. To produce high quality products, the critical
processing parameters must be controlled and the effect of disturbances minimized.
Following is a description of the variables used in process control.

Controlled Variables: these are the independent variables of a process and can be
defined as the parameters of the process indicating product quality or the operating
conditions of the process. It is also called output, a factor (variable) that is caused by a
system and which is used as a measure of performance for the given system. In a food
extrusion process (Figure 1), the controlled variables are the products quality attributes,
such as density, color, moisture content, etc.

Manipulated Variables: these are the dependent variables of the process used to cause
a change in the process. It is also called control input and is defined as a factor

©Encyclopedia of Life Support Systems (EOLSS)

FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

(variable) used to modify the system behavior. Therefore, inputs are variables that cause
or stimulate a change in system behavior. Extruder variables (Figure 1) that have a
significant effect on process variables or product attributes, and can be readily changed
or manipulated during processing, are considered manipulative variables for control
purposes. These manipulated variables include the screw speed, barrel temperature,
feed-rate, and water rate.

Disturbances: a signal that tends to adversely affect the value of the system’s output
and can be classified by internal or external disturbances. An internal disturbance is
referred to as those disturbances originating within the system, while an external
disturbance originates outside the system. In a food extruder (Figure 1), screw wear is
an internal disturbance and raw material variability is an example of external
disturbance to the system in control.

S
TE S
-

R
AP LS
-
C EO
TO ACCESS ALL THE 23 PAGES OF THIS CHAPTER,
Visit: http://www.eolss.net/Eolss-sampleAllChapter.aspx
E –

Bibliography
H
PL O

Bateson R.N. (1999). Introduction to Control System Technology. 711 pp. Upper Saddle River, NJ, USA:
Prentice Hall. [This presents the concepts and language of process control, including instrumentation and
communications].
M SC

Brescia L. and Moreira R.G. (1997). Modeling and control of a continuous frying process: I Dynamic
analysis and system identification. Trans. of Chem. Engr. 75(c), 3-11. [This work discusses continuous
frying principles and modeling].
SA NE

Brescia L. and Moreira R.G.. (1997). Modeling and control of a continuous frying process: II Control
development. Trans. of Chem. Engr. 75(c), 12-20. [This work shows the application of GPC controllers to
continuous frying systems].
U

Clarke D.W., Mohtadi C. and Tuffs P.S. (1987). Generalized predictive control - part I. The basic
algorithm. Automatica 23(2), 137-148. [This work provides the theory for the development of GPC
controllers].

Clarke D.W., Mohtadi C. and Tuffs P.S. (1987b). Generalized predictive control - part II. Extensions and
interpretations. Automatica 23(2), 149-160. [This work provides the fundamentals and applications of the
GPC controller].

Isermann R. (1989). Digital Control Systems. 659 pp. New York, USA: Spring-Verlag. [This presents a
comprehensive discussion on the fundamentals and applications of digital control, including
deterministic, stochastic, multivariable, and adaptive controls].

Moreira R. and Bakker-Arkema F.W. (1992). Grain dryers control: a review. Cereal Chemistry. 69(4),
390-396. [This work provides a review of the application of process control for continuous-flow grain
dryers].

©Encyclopedia of Life Support Systems (EOLSS)

FOOD ENGINEERING – Vol. IV - Process Instrumentation and Control - Moreira, Rosana G.

Schonauer S.L. (1995). Product quality adaptive control system for a food extruder. Ph.D. Dissertation,
Department of Agricultural Engineering, Texas A&M University, College Station, TX, USA. [This
presents a comprehensive study on food extrusion dynamics, modeling, process control development, and
implementation].

Seborg D.E., Edgar T.F. and Mellichamp D.A. (1989). Process Dynamics and Control. 716 pp. New
York: John Wiley & Sons. [This presents the fundamentals and the advanced control theory as applied to
chemical processes].

Biographical Sketch

Dr. Rosana G. Moreira was born in Brazil where she earned a Bachelor of Science in Agricultural
Engineering from Campinas State University in 1980. She received a Master of Science in 1992 and a
Doctor of Philosophy in 1989 in Agricultural Engineering from Michigan State University. Her
dissertation topic was in fundamental modeling and adaptive control of continuous flow grain dryers. Dr.
Moreira worked at Michigan State University for 6 months (1990) in the area of process control of food
extrusion processes as a post-doctor researcher. She joined the Department of Agricultural Engineering at

S
TE S
Texas A&M University in 1990, where she developed a research and teaching program in Food
Engineering. Currently, she is an Associate Professor and teaches courses in unit operations in food

R
AP LS
processing and food rheology. Her research emphasis is in fundamental modeling of food processing
systems (deep-fat frying, food dehydration, extrusion, etc.) and automatic control (predictive model
control) for food processing systems, such as food dehydrators, food extruders, and continuous fryers.
She has published more than 40 papers in professional journals, two book chapters, more than 50
C EO
abstracts and proceedings, and a book. Dr. Moreira is member of the American Society of Agricultural
Engineers and the Institute of Food Technologists and is a licensed Professional Engineer for Texas.
E –
H
PL O
M SC
SA NE
U

©Encyclopedia of Life Support Systems (EOLSS)