UNSTABLE = UNSAFE. WHAT IS THE COST OF STABILISATION OF AN UNSTABLE PLANT?

ŠEBEK, M.

Abstract

The field of Control engineering develops rapidly: numerous new control hardware and software appear every year. Similarly, new control theoretical methods are published helter-skelter. In contrast, basic control principles do not change: they held in the past, hold now and shall hold in the future as well. The basic principles cannot be changed or bypassed. They can only be forgotten but this may bring a lot of danger. Engineers control ever more interesting and more complex plants. The new plants are increasingly designed as unstable since it is profitable for their performance. The designers do it as they more and more believe in control engineers. But unstable plants are potentially dangerous. Even when stabilized, we must be cautious. If not, problems may come, remember Chernobyl. This talk aims to remember one basic control principle called Bode integral. Although it is often forgotten in the flood of modern and fashionable methods, the Bode integral restricts our possibilities very much, in particular when controlling an unstable plant. Such an unstable plant must be first stabilized but this costs something and hence we are as free as before to guarantee a desired performance.

Coresponding author e-mail: m[dot]sebek[at]c-a-k[dot]cz

Session: Plenary lecture