According to DIN 19226-4, the term "control" is to be based on a general structure. Control is thus the generic term for the totality of controlling and controlled system, similar to regulation, which includes both controller and controlled system. Figure 1 shows the control circuit.t.
In the controlling part of the system, the control algorithm processes information from various sources. Information from superimposed components and information from the user interface is included in the control. The process status is recorded via the sensors, and the status of the actuators is also fed back. For the control algorithm, this information is considered as input information and the output information is generated from it. The output information controls the actuators to influence the process to be controlled. In addition to the control commands to the actuators, the output information to the higher-level systems, but especially the information to the operator, are becoming increasingly important. Characteristic for such a control system is:
Input and output information of the controlling system is almost exclusively binary variables. Analogue input variables are mapped to binary variables and thus processed in the control algorithm.
Feedback from the process and actuators creates a control circuit.
Faults in actuators, processes and sensors must always be assumed, so fault processing is a fixed part of the control algorithm. Only foreseeable disturbances can be safely controlled, and in most cases, the complexity of the control algorithms only comes about through disturbance processing. A large part of the input signals is primarily or exclusively used for disturbance processing.
The processing of a large number of input and output signals in the control algorithm is typical. Even in simple industrial control solutions, fewer than 10 signals are rare, often several hundred and more. The task is to master this complexity and bring it into a functioning control algorithm without errors.
It is essential to add the perspective of the human operator or observer to the definition of the term "control". The output information from the controlling system must be comprehensible to humans and every state of the system must be recognisable to humans as operators or observers at all times.
The input information must always provide clear information to the superimposed systems and the human operator or observer through the control algorithm.
These requirements lead to the problem of control design.