Slaying Dragons #35 – Control Theory

John Watkinson ponders Control Theory.

One of the fundamentals of control theory is that when negative feedback is applied, the improvement in performance takes place at the point where the parameter being controlled is measured.

For example, if a loudspeaker is equipped with motional feedback that uses an accelerometer on the diaphragm, the accelerometer will faithfully follow the intended acceleration waveform because of the feedback. However, if the diaphragm is not sufficiently rigid, it will flex so that the reproduced sound waveform is not necessarily improved by the feedback.

Once we are familiar with that aspect of feedback, we can turn the idea on its head and use it to study systems that we either don’t understand or for which information is not available. For example, if we don’t understand what is being measured in a system using feedback, then we can come to some conclusions about that by considering what benefit eventually accrues.

These concepts are not only applicable to engineering problems. In fact engineering is easy, because the component parts tend to obey the laws of physics, which are true and constant, and analysis is usually possible. On the other hand once politics and/or the media enter the arena, the laws of physics are generally torn up and scattered and what we are told is what most people are prepared to believe, which should never be confused with the reality.

Whilst I remain committed to true science, knowledge that can be confirmed by repeatable experiments, I am afraid that performing true science is now something of a minority activity. If we look at the feedback loops that exist within academe, the most powerful one does not optimise the furtherance of knowledge. Instead it has become the seeking of funding and security of tenure.

Academic funding comes from politicians who are either totally ignorant of or who actively despise the tenets of science and prefer instead to cherry pick results that support their own dogma. In that context the freedom from bias that underlies real science cannot operate. A direct result is that…

(contd)
**To read the full article log in to the members’ site**