Approximate parameter invariance with nonlinear feedback

The principle of invariance, first discussed by Kulebakin [2] for systems subject to external disturbances, is extended to the problem of building systems that are insensitive to plant-parameter variations. Through the use of nonlinear signal comparators and plant models, several classes of systems that are relatively insensitive to large parameter variations have been devised. These feedback configurations implicitly identify major parameter changes and modulate the plant drive to compensate for these changes. The control of nonminimum phase plants and systems with noisy output transducers may be improved with these schemes.