This book considers the dynamics of a parallel kinematic production machine with hydraulic actuators. The maximum likelihood parameter estimation method is extended and proved effective with the model considered. Case studies apply these results and attain model paramaterizations that have a good fit to measured data. Model quality is further improved by use of experiments with greater information content. An optimization procedure is proposed that determines the experimental variables by balancing information content against experimental feasibility, while avoiding potential hazards. In practical comparisons, the use of the models resulted in experiments that had reduced controller action by over 60%. This in turn resulted in 50% lower tracking error. These improvements are due purely to updated model parameters, meaning that the extra effort spent on experiment design is rewarded. The novel aspect is that the experimental data underlying the parameter estimation was recorded at the machine in its ready-to-operate configuration.
|Publisher:||Stylus Pub Llc|
|Product dimensions:||5.80(w) x 8.10(h) x 0.80(d)|
About the Author
Franz Dietrich teaches at the Braunschweig University of Technology.