A general model of human/machine systems will be considered (see also van Cott and Kinkade, 1972; Ivergard, 1982).
Control is defined in international standards as a general concept to denote purposeful influence. Figure 2.8 shows schematically the main components of a control system. The system has input quantities (y) and output quantities (yo) respectively. The relationship between these input and output quantities is determined by a law or transfer function that is dependent on certain parameters (ps). The transfer function of a system may thus be expressed as an equation:
yo = ІЇУіР) (21
A system can be described if the inputs, the transfer function, the parameters, and the output quantities are known. Where one or more of these is unknown, various methods may be used to define them. The following combinations of known and unknown quantities can occur:
1. Inputs, transfer function, and parameters are known and the output is required, for example, in the evaluation of a system under design.
2. Outputs, functions, and parameters are known, and the inputs are required. The method used in this case is known as diagnostic and is used, for example, by doctors trying to find out the type of disease, i. e., trying to determine the reason (input) for the symptoms (output) shown by the patient.
3. Inputs, outputs, and transfer function are known, and the parameters are required. This method is known as identification, and is used, for example, when one wants a mathematical description of a particular event.
4. Inputs and outputs are known, and both the transfer function and the parameters are required. The method for this type of problem is called the ‘black box’ technique, and is the one commonly used in the description and testing of very complex technical systems such as computers.
In process industry control systems, all quantities are more or less well known, depending on how well ‘identified’ the system is; in other words, they are like option (3). A skilled operator or a well-developed computer control system ‘knows’ the various parameters well. In another case, one may be working with some form of ‘trial
 |
 |
 |
 |
 |
Disturbance
FIGuRE 2.8 Diagram of simplified control system.
and error’ philosophy. In practice there are always some unknown factors—disturbance quantities (y)—for which the operator/control system must compensate.
Two types of systems are concerned in control: the controlling system (control equipment) and the controlled system (process) (see Figure 2.8). The control equipment controls with the help of the controlling quantity, so that the desired value (set value) of the process can be maintained. The true value which the process returns is called the actual value. The process is also affected from outside, and this effect is known as the disturbance quantity.
