Engineering Mechanics

Proceedings Vol. 14 (2008)

ENGINEERING MECHANICS 2008

Copyright © 2008 Institute of Thermomechanics, Academy of Sciences of the Czech Republic, v.v.i., Prague

 

Krejčí P., Vlach R., Grepl R.
pages 128 - +8p., full text

The information about interaction between robotic parts and surroundings is necessary for intelligent control of robot behavior. This research is connected to research of contact force vector sensor principle and verification of its functionality which was performed in last year. Basic principle of sensor is based on measuring the strain in three locations by three strain gauges on active part of sensor. Based on these strains the contact force vector is identified by neural network. Huge matrix of training pairs is necessary for proper function of neural network. The training matrix contains pairs of strains in three locations and force vectors corresponding to sensor body strain. FE model of sensor was used to generate the training matrix. The research of this year was focused on further optimization of sensor body with ambition to improve its sensitivity for loads in all directions. The paper describes optimization procedure of sensor body as well as results of experimental verification of optimized sensor functionality.

Text and facts may be copied and used freely, but credit should be given to these Proceedings.

All papers were reviewed by members of the scientific committee.