Control Engineering
The goal of control engineering is to apply external influence to a dynamical system so that it behaves in a desired way. Here, the behavior is under constant monitoring through a comparison between the desired and the actual measured values of the system
- PDF / 788,773 Bytes
- 35 Pages / 439.36 x 666.15 pts Page_size
- 70 Downloads / 210 Views
Control Engineering
The goal of control engineering is to apply external influence to a dynamical system so that it behaves in a desired way. Here, the behavior is under constant monitoring through a comparison between the desired and the actual measured values of the system. When these deviate, the goal is to manipulate the system in a way to minimize future deviations. In view of this feedback one speaks of closed-loop control or feedback control. If there is no feedback, i.e., if the manipulation happens without knowledge on the actual state of the system, then one does not speak of a closed-loop controller but instead of an open-loop or non-feedback control. In everyday life, the human himself very often acts as a controller, e.g., in the shower when adjusting the water temperature. The desired state is a comfortable water temperature while unfortunately the actual state strongly deviates much too often. We keep changing the setting of the faucet until the deviation is within our threshhold of acceptability. There appear controls in nature as well. In our body there exist a multitude of control loops which keep quantities such as the body temperature or the blood pressure at certain values, for example. Nowadays, control engineering finds its predominant application in technical systems. Well-known application areas include air conditioning systems or systems that enhance driving safety such as ABS and ESC. But control engineering also finds applications in non-technical systems, for example in economics or in the analysis of biological controls. In any case, the modeling of the system that is to be controlled is always a part of control engineering. In Sect. 11.2, we will provide examples whose purpose is to provide details of such modeling by means of a mechanical system. There exist many different approaches that can be tapped to reach a given control objective. Methods in the area of “classical control theory”, which we will briefly cover in Sect. 11.1.4, require that a controller is developed in such a way that the closed control loop displays certain mathematically definable properties. One advantage of this procedure is that one can easily precompute the reaction of this control loop to certain inputs which therefore allows the ability to satisfy quite diverse quality criteria through a suitable design of the controller. H.-J. Bungartz et al., Modeling and Simulation, Springer Undergraduate Texts in Mathematics and Technology, DOI 10.1007/978-3-642-39524-6__11, © Springer-Verlag Berlin Heidelberg 2014
255
256
11 Control Engineering
Fig. 11.1 Simple mechanical controller to control inflow
A problem with classical control theory, however, is the fact that a lot of information about the controlled system has to be known in order to set up a reasonably reliable model of the system. Furthermore, the subsequent design of a controller is often very complicated. This effort may not be necessary for all cases. In fuzzy control, which accounts for the main part of this chapter, the design of a controller at
Data Loading...
