Process control, PID and adaptive control - Intelligent Buildings and Building Automation - page 112

Process disturbance
inputs

Disturbance
compensation

Disturbance
sensors

Desired value
of controll ed
variable

Controlled
vari able

Actuation
device

Controller Process

Figure 7.2 Disturbance compensated open-loop system.

signal for the actuation devices by implementing the control law or control
algorithm, which is practically based on the predicted correlation between
the process input and output.
Open-loop systems, without considering the effects of disturbances, as
shown in Figure 7.1, can be satisfactory or acceptable if the disturbances are
not great and the changes in desired values are not too severe. However, in
many control systems in practice, the effects of disturbances are great or the
errors in the controlled variables caused by such disturbances are serious. To
reduce errors in the controlled variable in the open-loop system, the effects
of the disturbances should be considered by the control scheme in generating
the control signals. Implementation of such a scheme requires, as shown in
Figure 7.2, that:

￿ the disturbances can be measured;
￿

the effects of the disturbances on the controlled valuable can be estimated,
so that they can be compensated for.

It can be very costly and sometimes impractical to quantify the disturbances
and to estimate the effects accurately. We can use the room temperature
control in a heated room as an example. The heating load of the room is
strongly affected by the outdoor air temperature, solar radiation, infiltration,
internal heat from people, lighting and equipment, and so on. To control the
output of the heat and therefore control the room temperature accurately,
first, we need to measure the changes in the external and internal disturbance
factors. Second, we should have an accurate mathematical model correlate
these external and internal parameters on the heating load. When accurate
control of the room air temperature is needed, the control system needed
will be very complex, if not impossible, even for such a simple application.
Certainly, open-loop systems are stable and have no control stability prob-
lems as occur in closed-loop systems.
To solve the problem of complexity, a simple solution can be very effective.
That is to use the controlled variable, the final target of the system, as the
reference in making the control decision. Let us consider how an operator
can control the room temperature in this case.