The static and dynamic performance of a control system depends on the accuracy of the mathematical model of the plant that is being controlled. In this work, the accuracies of a linear and a second-order kinematic model were evaluated for a two-dimensional, planar, micro-positioner plant. The model performances were evaluated for a variety of testing conditions including a variation of the number of test points, the addition of sensor noise, and six different least-squares fitting algorithms. In general, decreasing the number of fitting points or increasing the noise results in worse performance, as expected. The second-order model was better for low noise and larger data sets, but was worse under the more challenging conditions. All but one of the least-squares algorithms performed similarly well.
Conference Dates: November 10-15, 2001
Conference Location: Arlington, VA
Conference Title: Proceedings of the American Society for Precision Engineering
Pub Type: Conferences
Kinematic, Micro-positioner, Model, Performance