Feedback System Theory.
Abstract
The basic reason for using feedback in control is to cope with uncertainty in plant parameters and plant disturbances, and so obtain plant performance otherwise unattainable. However, most of the feedback control literature has tended to be qualitative rather than quantitative, with uncertainties undefined or even completely ignored. This research has been highly quantitative with great emphasis on precise definition of the plant uncertainty and of the performance tolerances, to be achieved despite the uncertainty. Also, the practical aspects have been considered by attention to the gain and bandwidth of the compensation needed in the feedback loops. This is the true measure of the cost of feedback, which as been ignored in 'modern' state-variable control theory, leading them to totally impractical unachievable infinite bandwidth amplifiers. In our research in quantitative synthesis, the uncertainties are precisely defined, the compensation must be realizable, and practical, transparent synthesis techniques are sought. A transparent technique is one in which the designer clearly sees, as he proceeds, the trade-offs between the important design factors, and can choose between them; for example, bandwidth and therefore sensor noise effect conservation versus compensation complexity. During the five-year period, quantitative, transparent synthesis techniques have been developed for the first time, for certain classes of feedback control systems. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1980
- Accession Number
- ADA100204
Entities
People
- Isaac M. Horowitz
Organizations
- University of Colorado Boulder