A remote laboratory on PID autotuninghttp://www.nt.ntnu.no/users/skoge/prost ... s/0532.pdfAbstract: This manuscript presents a remote, web-enabled laboratory devoted to PID autotuning—a subject of significant importance from both the theoretical and the application-oriented point of view, but seldom available in remote laboratories. In detail, the manuscript presents two PI/PID autotuning experiments on physical control systems; the experiments are accessible by means of a browser, and the code is available as free software. Some pedagogical issues are also briefly discussed.
Цитата:
A huge number of automatic control experiments is nowadays offered by remote laboratories, as witnessed by works such as Aktan et al. (1996); Valera et al. (2005); Aliane et al. (2006), and many others; such resources are nowadays very important for control education at all levels (Dormido Bencomo, 2004).
However, a very small minority of experiments available from remote laboratories is relative to the domain of autotuning, a
notable exception being e.g. Balda et al. (2004).
This is a lack to be filled, for at least three reasons. First, there is a vast literature on autotuning, with particular reference to the PI/PID controller structure (A° stro¨m and Ha¨gglund, 1995; Leva et al., 2002; O’Dwyer, 2003). When a subject is so extensively
investigated, also the availability of well designed laboratories on that subject becomes a necessity, both for research and for education purposes. Second, autotuning capabilities are now included in many industrial regulators, see e.g. Li et al. (february 2006), and users tend more and more to rely on them. This fact is of particular importance in process control, for example, where a systematic application of autotuning to the hundreds of low- or mid- level loops involved in the typical plant allows to shorten commissioning and maintenance time significantly, and also to focus attention on the top-level plant controls. Also in motion control, to quote a somehow complementary field with respect to process control, autotuning yields significant benefits, as witnessed by the numerous products that nowadays offer such a feature. In synthesis, then, mastering not only the practical the use, but also (which is even more important) the
theory of operation of an autotuner is important for the control engineer—and is not as easy a task as one may think at a first glance, as will appear later on. Third, constructing an autotuner is a challenging and fascinating activity, and probably the core of the competencies to induce when educating engineers. Turning a method to (auto)tune a regulator from the form it takes in the scientific literature to that of a complete procedure capable of running autonomously on a control computer involves a large amount of design choices, which can be taken consciously only if the designer possesses an adequate mix of control-theoretical and software engineering capabilities.