Nonlinear Identification and Controller Design for the Air Levitation System

Abstract

This paper presents an alternative method for modeling the air levitation system using nonlinear identification, to overcome common mismatches between simulation and the real system. The proposed controller pursues to compensate the nonlinearities present in the system, allowing the design of classical PID controller. Within the results are presented the behavioral comparison between a nonlinear and a linear identified system. The presented approach aims to use available simulation tools, so that the results can be included as part of laboratory experiments.