Training Possibilities in CPN

  • V. Baranauskas Kaunas University of Technology
  • K. Šarkauskas Kaunas University of Technology
  • S. Bartkevičius Kaunas University of Technology

Abstract

Flexible production systems require usage of mobile transportation robots. Training of these robots is necessary to ensure effective and conflict less operation. The principles of multi-agent systems and artificial neural nets may be used in training processes, but some not solved yet problems make usage of these methods complicated. Colored Petri nets are well suited for purposes of training, if exten-sion of usage of global variables made. Presence of these variables creates the possibility to accumulate data obtained during training and provide databases. The training of system is conducted on the model; collisions are resolved using graphical subsystem and data-base, accessible for all mobile robots, provided. The database of the already trained system of control can be transferred to the control device. Ill. 4, bibl. 7 (in English; summaries in English, Russian and Lithuanian).

Published
2007-04-19
How to Cite
Baranauskas, V., Šarkauskas, K., & Bartkevičius, S. (2007). Training Possibilities in CPN. Elektronika Ir Elektrotechnika, 76(4), 25-28. Retrieved from http://eejournal.ktu.lt/index.php/elt/article/view/10710
Section
T 125 AUTOMATION, ROBOTICS