Multivariable control of a magnetic levitation system using closed loop identification and H/sub ∞/ control theory

Time-domain fixed-structure closed-loop model identification of an unstable multivariable maglev nanopositioning system

“…System identification is required to validate the analytical model (8)- (10) and to identify the dynamic coupling among the axes due to the non-ideal actuators and mechanical structure of the positioner, and misalignments in its assembly. Open-loop tests cannot be performed due to the inherently unstable nature of the maglev system [2].…”

“…8 demonstrates the effect of the identified model on better and more accurate controller design. This figure shows the plant input and closed-loop responses to a 100-µm reference step of the actual controlled system with controllers designed using identified model (22) and analytical model (10). It also shows the simulated response using identified plant TF models.…”

“…Unlike the decoupled control with the linearized plant TFs about an operating point as given by (10), this MIMO control uses outputs from all three modes and calculates the controller input for the plant as well as the coupling TFs. The closed-loop system response with the designed controller is therefore expected to meet the objectives outlined in Section 1.…”

“…Lin and Jou [9] developed an improved force-model-identification method for magnetic suspension systems to establish reliable parameters to describe the nonlinear current-todistance characteristics of the magnetic field. Nakashima et al [10] identified the vibration characteristics of a three-input/output maglev system using closed-loop identification, and designed an H ∞ servo controller which can effectively suppress these vibrations. A direct closed-loop identification method using the plant input/output data acquired through an output intersampling scheme was presented, and its effectiveness was demonstrated through an experimental study using a magnetic-suspension system by Sun et al [4].…”

APPLICATION OF H∞ CONTROL AND CLOSED LOOP IDENTIFICATION TO A MAGNETIC LEVITATION SYSTEM

“…In view of the academic background of control applications, a special issue on magnetic bearing control [3] has been published. We have been researching a multivariable control of a magnetic levitation system with a Y shape iron plate as a basic study of such mechanisms [4][5][6][7][8][9]. This system has three major obstacles to its stabilization such as input-output coupling, nonlinearities of electromagnetic forces, and destabilization of the closed loop system due to spillovers.…”