The unwanted cogging forces are caused by the magnetic attraction between the iron core in the motor’s primary and the permanent magnets in the secondary. The new software takes into account not only these magnetic effects, but also those of energy chains and the application’s mechanical design. As a result, it expands the range of potential applications for iron-core linear motors “significantly”, according to Beckhoff.
The compensation tool is based on the automated application of machine learning in Beckhoff’s TwinCat software and is designed to be used with the company’s AL8000 linear servomotors. The software records the cogging data in an application during a reference run over the entire length of the motor’s magnetic track. Using this data, it trains a neural network, which is integrated with the control system for current pre-control. By adapting the current pre-control in this way, the lag error can be reduced by a factor of up to seven, while the synchronisation of the machine increased up to five-fold without needing any hardware changes to the linear motor.
Beckhoff’s cogging compensation software is designed for use with its AL8000 linear servomotors
Beckhoff Automation: Twitter LinkedIn Facebook
