Intelligent TDA Systems Using Neural Networks Phase 1 Final Report

The overall objective of Phase I was to develop a prototype intelligent Tactical Decision Aids TDA neural network system through the use of an expert system to neural network translator. From the resulting prototype network, the performance of the complete TDA system could be determined. The Phase I results absolutely demonstrated the feasibility of this concept and cleared the way for development of a complete system in Phase II. Current TDA systems require too much expertise from the user and run too slowly for use in tactical situations, when time is in short supply. In tactical situations, terrain is an extremely important factor and represents one of the main inputs to these systems. The Airborne Avenues of Approach AAA Planning problem was chosen. While initial route planning is performed prior to the start of the mission, a need exists to automate and speed the process. Additionally, a system which could meet the rigorous real-time requirements of an on-board AAA re-planner would prove enormously beneficial in many time-critical tactical situations.