M. Feemster

Tracking control of mechanical systems in the presence of nonlinear dynamic friction effects

In this paper, we design an observer-based exact model knowledge position tracking controller for a second-order mechanical system with nonlinear load dynamics and a nonlinear dynamic friction model. Since the controller requires an estimate of the unmeasurable friction state, we demonstrate how the friction dynamics can be exploited to design three different observers which foster different transient response characteristics for the composite closed-loop system. We then present two adaptive controllers which utilize nonlinear observer/filter structures to provide for asymptotic position tracking while compensating for selected parametric uncertainty. Dynamic simulation and experimental results are utilized to illustrate control performance.

Sensorless rotor velocity tracking control for induction motors

We present a sensorless control algorithm that achieves semi-global exponential rotor velocity tracking for the full-order nonlinear dynamic model of an induction motor actuating a mechanical subsystem. The proposed controller utilizes stator current measurements, but is termed sensorless due to the fact that no mechanical sensors are required and that stator current measurements can be obtained in an inexpensive/simplistic manner. The control strategy utilizes a novel rotor velocity observer which facilitates the potential for improved rotor velocity tracking transient performance. In addition, the observed integrator backstepping technique is utilized to ensure that the observer-based controller remains bounded. Experimental results are included to verify the effectiveness of the controller.

Adaptive control techniques forfrictioncompensation

Abstract In this paper, we design two adaptive controllers for a second-ordermechanicalsystem which incorporates frictional effects such as Coulomb, static, Stribeck, andviscousfriction. First, we design a modular position tracking controller that can accommodate avarietyof adaptive update laws. The proposed controller is shown to compensate foruncertaintyassociated with the friction parameters which appear linearly in the model. In thesecond controlscheme, we show how a Lyapunov-based adaptive position setpoint controller canbe designed tocompensate for parametric uncertainty throughout the mechanical systemincluding the Stribeckeffect related constant which does not appear linearly in the model.Experimental results areprovided to illustrate the performance of the proposed controllers.

Adaptive control techniques for friction compensation

We design two adaptive controllers for a second-order mechanical system which incorporates frictional effects such as Coulomb, static, Stribeck, and viscous friction. First, we design a modular position tracking controller that can accommodate a variety of adaptive update laws. The proposed controller is shown to compensate for uncertainty associated with the friction parameters which appear linearly in the model. In the second control scheme, we show how a Lyapunov-based adaptive position setpoint controller can be designed to compensate for parametric uncertainty throughout the mechanical system including the Stribeck effect related constant which does not appear linearly in the model. Experimental results are provided to illustrate the performance of the proposed controllers.

Vision Assisted Autonomous Landing of an Unmanned Aerial Vehicle.

In this paper, a strategy for an autonomous landing maneuver for an underactuated, unmanned aerial vehicle (UAV) using position information obtained from a single monocular on-board camera is presented. Although the UAV is underactuated in translational control inputs (i.e., a lift force can only be produced), the proposed controller is shown to achieve globally uniform ultimate boundedness (GUUB) in position regulation error during the landing approach. The proposed vision-based control algorithm is built upon homography-based techniques and Lyapunov design methods.

Nonlinear control techniques for an atomic force microscope system

Two nonlinear control techniques are proposed for an atomic force microscope system. Initially, a learning-based control algorithm is developed for the microcantilever-sample system that achieves asymptotic cantilever tip tracking for periodic trajectories.Specifically,the control approach utilizes a learning-based feedforward term to compensate for periodic dynamics and high-gain terms to account for non-periodic dynamics. An adaptive control algorithm is then developed to achieve asymptotic cantilever tip tracking for bounded tip trajectories despite uncertainty throughout the system parameters. Simulation results are provided to illustrate the efficacy and performance of the control strategies.

An improved indirect field-oriented controller for the induction motor

In this paper, the standard indirect field-oriented controller (IFOC) commonly used in current-fed induction motor drives is modified to achieve global exponential rotor velocity/rotor flux tracking. The modifications to the IFOC scheme, which involve the injection of nonlinear terms into the current control input and the so-called desired rotor flux angle dynamics, facilitate the construction of a standard Lyapunov stability argument. The construction of a standard Lyapunov exponential stability argument allows one to easily design adaptive controllers to compensate for parametric uncertainty associated with the mechanical load. Simulation results are included to illustrate the improvement in performance over the standard IFOC scheme.

Sensorless rotor velocity tracking control of the permanent magnet stepper motor

We present a sensorless rotor velocity tracking controller for the full order, nonlinear dynamic model of the permanent magnet stepper motor actuating a mechanical subsystem. Specifically, the structure of the electrical subsystem dynamics is exploited to reconstruct the rotor position and velocity signals from measurements of stator current and stator voltage. This surrogate rotor position signal is then used to design a control strategy that achieves exponential rotor velocity tracking. Experimental results are included to demonstrate the efficacy of the proposed algorithm.

Non-linear adaptive control of induction motors

Given measurements of rotor position, rotor velocity, and stator currents, we design an adaptive control scheme that is free of singularities, does not require rotor flux measurements, and provides for simultaneous asymptotic rotor position/rotor flux tracking despite the uncertainty associated with the mechanical subsystem parameters and the rotor resistance parameter. For the case when the rotor resistance parameter is known exactly, we modify the structure of the controller to achieve global asymptotic rotor position/rotor flux tracking while accommodating for parameter uncertainty associated with the mechanical subsystem parameters and the stator electrical subsystem parameters. Experimental results are presented to illustrate the performance of the control structure. Copyright

Active interaction force identification for atomic force microscope applications

In an effort to improve sub-molecular imaging precision, a general distributed-based modeling approach for the atomic force microscope system is proposed which allows for the design of a stabilizing controller/estimation scheme that stabilizes the micro-cantilever based system while asymptotically identifying the atomic interaction force, that is, the estimated interaction force asymptotically approaches the actual interaction force; hence the estimated interaction force can be utilized to generate high precision atomic-resolution images. Differing from current practices of simple lumped model sets of ordinary differential equations, the proposed approach attacks the more difficult distributed parameter model.