RESEARCH IN AI FOR ON-BOARD MISSION/PAYLOAD MANAGEMENT IN MULTI PAYLOAD UNMANNED AIR VEHICLES

THE FEASIBILITY OF AN ARTIFICIAL INTELLIGENCE SYSTEM FOR UNMANNED AIR VEHICLE (UAV) MISSION PLANNING, NAVIGATION WAYPOINT SELECTION AND INTEGRATED ESM/EO PAYLOAD OPERATIONS WILL BE DEMONSTRATED VIA RAPID PROTOTYPING. THE MISSION PLANNING ALGORITHM WILL RECOGNIZE CONFLICTS AMONG MILITARY/OPERATIONAL REQUIREMENTS WHICH ARE COMPETING FOR AVAILABLE UAV RESOURCES; A SYSTEM OF RULES WILL BE APPLIED TO RECONCILE CONFLICTS. THE NAVIGATION WAYPOINT SELECTION ALGORITHM WILL SELECT THE MOST EFFICIENT FLIGHT PATH FOR IMPLEMENTING THE MISSION PLAN SUBJECT TO RULES FOR MINIMIZING VULNERABILITY OF ENEMY AIR DEFENSE. THE ESM/EO PAYLOAD OPERATIONS MODULE WILL: 1) CORRELATE EMITTER LINES OF BEARINGS ACQUIRED BY THE ESM PAYLOAD AGAINST REDEPLOYABLE TARGETS AT DIFFERENT TIMES, 2) CORSS-CORRELATE DIFFERENT EMITTERS WHICH ARE ASSOCIATED WITH C3 NODES AND COMBAT UNITS TO FORM A MATRIX OF RADIO NETWORKS AND NODE RELATIONSHIPS, 3) APPLY A "MINIMUM ENTROPY" ALGORITHM TO DETERMINE THE MINIMUM SET OF NODES WHICH SHOUD BE IMAGED BY THE EO PAYLOAD IN ORDER TO RESOLVE AMBIGUITIES IN THE NETWORK/NODE MATRIX, AND 4) APPLY AI PATTERN RECOGNITION TECHNIQUES TO FUNCTIONALLY IDENTIFY NETWORKS AND NODES. RAPID PROTOTYPING WILL EMPLOY NETTED COMPAQ COMPUTERS AND HIGH-LEVEL LANGUAGE TO DEMONSTRATE FEASIBILITY OF ALGORITHMS.