SELECTED TYPES OF NEURAL NETWORKS FOR MAGNETOELASTIC SENSOR ERROR SUPPRESSION

Jozef Vojtko

   Magnetoelastic sensors belong to non-linear systems with hysteresis. These sensors used in measurement are prone to errors due to their material properties. The sensor errors can be suppressed by means of conventional or unconventional techniques - neural networks. The paper brings out the results of magnetoelastic sensors errors suppression by using selected types of neural networks.

Keywords: measurement, magnetoelastic sensor, neural network, non-linearity, hyteresis

[full-paper]