Compensation for Unbalances with Aid of Neural Networks
The aim of this paper is to provide an extended method to compensate remaining unbalances at magnetic bearings. The compensation signal will still be generated by a MLP-network. Only now, the network needs very few training data, other values will be learned during operation with the aid of a quality criterion.
Author: M. Paul and W. Hofmann and H. F. Steffani | Published: 1998
Booktitle: Proceedings of ISMB6
Booktitle: Proceedings of ISMB6