Title :
Analysis of constraint satisfaction neural nets for route planning
Author :
Welker, Kevin B. ; Barhorst, James F.
Author_Institution :
McDonnell Douglas Missile Syst. Co., St. Louis, MO, USA
Abstract :
Constraint satisfaction neural network simulations were used to perform route planning. Three types of models were used, schema, interactive activation competition, and Boltzmann machine, each with unique characteristics. The performances of the three models in performing routing functions were compared and analyzed, and significant factors stemming from their performance properties were identified for the exhibited differences between them
Keywords :
digital simulation; military computing; neural nets; performance evaluation; planning (artificial intelligence); Boltzmann machine; constraint satisfaction neural nets; deterministic models; interactive activation competition; route planning; schema model; tactical data processing; threats; weighting; Artificial intelligence; Artificial neural networks; Biological neural networks; Computer networks; Constraint optimization; Missiles; Network topology; Neural networks; Routing; Steady-state;
Conference_Titel :
Digital Avionics Systems Conference, 1991. Proceedings., IEEE/AIAA 10th
Conference_Location :
Los Angeles, CA
DOI :
10.1109/DASC.1991.177209
