In a paper in the context of controller design for DEAP actuators I came upon the following sentence:

Furthermore, the nonlinear behavior of the control plant requires either an adaptive or a robust design in order to ensure a precise and accurate control.

What is the relationship between a robust/adaptive controller and the nonlinearity of the control plant? Isn't it somehow always necessary to design a robust controller no matter whether the control plant is linear or not? Or are nonlinear plants in general more sensitive to parameter variations, e.g. nonlinear systems sensitive to initial conditions (chaotic behaviour)?


The authors are assuming that you're going to make a linear or mostly linear controller after finding a linearized model of the plant.

It's not that nonlinear plants are sensitive to model variations, it's that for a certain class of nonlinear plant you can treat the plant as a linear time-varying system that has time-varying disturbances.

So in that sort of a design paradigm, you use some design process that works with time-varying plants: either you make an adaptive controller that's constantly discovering the up-to-date plant model and controlling it in some optimal way, or you make a robust controller that will control a linear time-varying plant that varies in the way your linearization does over the range of operating points of the real, nonlinear plant.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.