A systematic, rigorous, and practice oriented method for control system performance evaluation is proposed. The approach is rigorous by employing techniques from statistics like regression analysis, optimal design of experiments, and hypothesis testing. The method is implemented with tools that are readily applicable. The tools were applied on an experimental rig evaluating four controllers. Statistically significant differences between the performance of the control schemes were found, while it did not appear that the more complicated control schemes consistently perform better.
|Title of host publication||Proceedings the 4th international conference on motion and vibration control, [August 25-28, 1998, ETH Zurich, Switzerland]. Vol. 3|
|Editors||G. Schweitzer, R. Siegwart, P. Cattin|
|Place of Publication||Zurich|
|Publisher||Institute of Robotics, ETH|
|Publication status||Published - 1998|