Process Control 过程控制

Lesson 4 Process Control

1 Process-Control Principles

Process Control is the active changing of the process based on the results of process

monitoring. In process control, the basic objective is to regular the value of some quantity. To

regulate means to maintain that quantity at some desired value regardless of external influences.

The desired value is called the reference value or setpoint.

Figure 5.9 shows the process to be used for this discussion. Liquid is flowing into a tank at

some rate Qin and out of the tank at some rate Qout. The liquid in the tank has some height or

level h. It is known that the flow rate out varies as the square root of the height, so the higher the

level the faster the liquid flows out. If the output flow rate is not exactly equal to the input flow

rate, the tank will either empty, if Qin > Qout, or overflow, if Qout < Qin.

This process has a property called self-regulation. This means that for some input flow rate,

the liquid height will rise until it reaches a height for which the output flow rate matches the input

flow rate. A self-regulating system does nit provide regulation of a variable to any particular

reference value. In this example the liquid level will adopt some value foe which input and output

flow rates are the same and there it will stay. If the input flow rate changed, then the level would

change also, so it is not regulated to a reference value.

Suppose we want to maintain the level at some particular value H in Figure 5.9, regardless of

the input flow rate, then something more than self-regulation is needed.

Human-Aided Control

Human-aided control shows a modification of the tank system to allow artificial regulation of

the level by a human. To regulate the level so that in maintains the value H it will be necessary to

employ a sensor to measure the level. This has been provided via a sight tube. The actual liquid

level or height is called the controlled variable. In addition, a value has been added so the output

flow rate can be changed by the human. The output flow rate is called the manipulated variable or

controlling variable.

Now the height can be regulated apart from the input flow rate using the following strategy:

The person measures the height in the sight tube and compares the value to the setpoint. If the

measured value is larger, the human opens the value a little to let the flow out increase, and thus

the level lowers toward the setpoint. If the measured value is smaller than the setpoint, the person

closes the value a little to decrease the flow out and allow the level to rise toward the setpoint.

By a succession of incremental opening and closing of the value, the human can bring the

level to the setpoint value H and maintain it there by continuous monitoring of the sight tube and

adjustment of the value.

Automatic Control

To provide automatic control, the system is modified using machines, electronics, or

computers to replace the operations of the human. An instrument called a sensor is added that is

able to measure the value of the level and convert it into a proportional signal. The signal is

provided as input to a machine, an electronic circuit, or a computer, called the controller. This

performs the function of the human in evaluating the measurement and providing an output signal

to change the value setting via an actuator connected to the valve by a mechanical linkage.

2 Identification of Elements

The elements of a process-control system are defined in terms of separate functional parts of

the system.

Process

In general, a process can consist of a complex assembly of phenomena that relate to some

manufacturing sequence. Many variables may be involved in such a process, and it may be

desirable to control all these variables at the same time. There are single-variable process, in

which only one variable is to be controlled, as well as multivariable processes, in which many

variables, perhaps interrelated, may require regulation.

Measurement

Clearly, to effect control of a variable in a process, we must have information on the variable

itself. Such information is found by measuring the variable. In general, a measurement refers to

the conversion of the variable into some corresponding analog of the variable, such as a pneumatic

pressure, an electrical voltage, or a current. A sensor is a device that performs the initial

measurement and energy conversion of a variable into analogous electrical or pneumatic

information. Further transformation or signal conditioning may be required to complete the

measurement function. The result of the measurement is a representation of the value in some

forms required by the other elements in the process-control operation.

Controller

The next step in the process-control sequence is to examine the error and determine what

action, if any, should be taken. This part of the control system has many names; however,

controller is the most common. The evaluation may be performed by an operator, by electronic

signal processing, by pneumatic signal processing, or by a computer. The controller requires an

input of both a measured indication of the controlled variable and a representation of the reference

value of the variable, expressed in the same terms as the measured value.

Control Element

The final element in the process-control operation is the device that exerts a direct influence

on the process; that is, it provides those required changes in the controlled variable to bring it to

the setpoint. This element accepts an input from the controller, which is then transformed into

some proportional operation performed on the process.

3 Process-Control Block Diagram

To provide a practical, working description of process control, it is useful to describe the

elements and operations involved in more generic terms. Such a description should be independent

of a particular application and thus be applicable to all control situations. A model may be

constructed using blocks to represent each distinctive element. The characteristics of control

operation then may be developed from a consideration of the properties and interfacing of these

elements. Figure 5.10 shows a general block diagram. The controlled variable in the process is

denoted by e in this diagram, and the measured representation of the controlled variable is labeled

b. The controlled variable setpoint is labeled r, for reference.

The error detector is a subtracting-summing point that outputs an error signal e=r-b to the

controller for comparison and action.

The purpose of a block diagram approach is to allow the process-control system to be

analyzed as the interaction of smaller and simpler subsystems. If the characteristics of each

element of the system can be determined, then the characteristics of the assembled system can be

established by an analytical marriage of these subsystems. The historical development of the

system approach in technology is dictated by this practical aspect: first, to specify the

characteristics desired of a total system and, then, to delegate the development of subsystems that

provide the overall criteria..

4 Control System Evaluation

A process-control systems is used to regulate the value of some process variable. When such

a system is in use, it is natural to ask, “How well is it working?” This is not an easy question to

answer, because it is possible to adjust a control system to provide different kinds of response to

errors. This section discusses some methods for evaluating how well the system is working.

The variable used to measure the performance of the control system is the error, which is the

difference between the constant setpoint or reference value r and the controlled variable c(t).

Since the value of the controlled variable may vary in time, so may the error.

In principle, the objective of a control system is to make the error exactly zero, but the

control system responds only to errors. Conversely, if the error were zero and stayed zero, the

control system would be doing nothing and would not be needed in the first place. Therefore, this

objective can never be perfectly achieved, and there will always be some errors. The question of

evaluation becomes one of how large the error is and how it varies in time.

The purpose of the control system is to regulate the value of some variables. This requires

that action be taken on the purpose itself in response to a measurement of the variable. If this is

not done correctly, the control system can cause the process to become unstable. In fact, the more

tightly we try to control the variable, the greater the possibility of an instability.

The first objective, then, simply means that the control system must be designed and adjusted

so the system is stable. Typically, as the control system is adjusted to give better control, the

likelihood of instability also increases.