Research Papers

Output Saturation in Electric Motor Systems: Identification and Controller Design

[+] Author and Article Information
Kyoungchul Kong1

Department of Mechanical Engineering, University of California, Berkeley, Berkeley, CA 94720kckong@me.berkeley.edu

Helge C. Kniep

Department of Mechanics and Ocean Engineering, Technische Universität, Hamburg-Harburg 94720, Germanyhelge.kniep@tuhh.de

Masayoshi Tomizuka

Department of Mechanical Engineering, University of California, Berkeley, Berkeley, CA 94720tomizuka@me.berkeley.edu


Corresponding author.

J. Dyn. Sys., Meas., Control 132(5), 051002 (Aug 11, 2010) (8 pages) doi:10.1115/1.4001792 History: Received April 23, 2009; Revised April 29, 2010; Published August 11, 2010; Online August 11, 2010

Input saturation is a well-known nonlinearity in mechanical control systems; it constrains the maximum acceleration, which results in the limitation of the system response time. Input saturation has been considered in controller design in various ways, e.g., antiwindup control. In addition to the input, the state variables of mechanical systems are often subjected to saturation. For example, the maximum angular velocity of electric motor systems is limited by the maximum voltage provided to the motor windings. In the case of electronically commutated motors (i.e., brushless dc motors), the maximum speed is additionally constrained by limitations of the servo amplifier output. If gears are utilized, further constraints are introduced due to resonances in ball bearings and/or velocity dependent friction. Although such factors are significant in practice, they have not been fully considered in controller design. This paper investigates the input and output saturations, and presents how they may be considered in the controller design; a Kalman filter, a PID controller, and a disturbance observer are designed, taking input/output saturations into consideration. A case study is provided to verify the proposed methods.

Copyright © 2010 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

The geared BLDC motor system used as the experimental testbed

Grahic Jump Location
Figure 2

Frequency responses of the system showing the system nonlinearities in the lower frequency range

Grahic Jump Location
Figure 3

Effects of the system velocity saturation on the system response

Grahic Jump Location
Figure 4

Velocity responses to step inputs with different magnitudes

Grahic Jump Location
Figure 5

Velocity dependant forces; the system shows an offset value due to coulomb friction as well as velocity dependent friction components

Grahic Jump Location
Figure 6

Hybrid system model considering velocity saturation

Grahic Jump Location
Figure 7

System model with input/output saturations: (a) the input saturation of the system, (b) the output saturation, and (c) the Coulomb friction

Grahic Jump Location
Figure 8

Simulations for model validation; the response of the hybrid system model closely matches with the experimentally obtained data: (a) input sequence I: large magnitude and low frequencies; (b) input sequence II: small magnitude and high frequencies

Grahic Jump Location
Figure 9

Kalman filter structure; the hybrid system is used as basis for the Kalman filter

Grahic Jump Location
Figure 10

Kalman filter performance; the effects of input and output saturations in the Kalman filter design are demonstrated: (a) position estimation; (b) velocity estimation

Grahic Jump Location
Figure 11

Disturbance observer considering input/output saturations

Grahic Jump Location
Figure 12

Estimated disturbances of different DOBs; the effect of input/output saturations in DOB systems is demonstrated: (a) estimated disturbance neglecting input and output saturation; (b) estimated disturbance neglecting the input saturation; (c) estimated disturbance neglecting the output saturation; and(d) estimated disturbance considering both input and output saturation

Grahic Jump Location
Figure 13

PID control with output antiwindup loop; the integral action is suppressed when the system velocity reaches its saturation value

Grahic Jump Location
Figure 14

Performance of PID controllers with and without output antiwindup; the output antiwindup loop in the controller prevents system overshoots




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In