ABDUCTIVE/EXPERT SYSTEM CONTROLLER

Award Information
Agency: Department of Defense
Branch: Army
Contract: N/A
Agency Tracking Number: 18452
Amount: $250,000.00
Phase: Phase II
Program: SBIR
Awards Year: 1994
Solitcitation Year: N/A
Solitcitation Topic Code: N/A
Solitcitation Number: N/A
Small Business Information
Abtech Corp
700 Harris Street, Charlottesville, VA, 22901
Duns: N/A
Hubzone Owned: N
Woman Owned: N
Socially and Economically Disadvantaged: N
Principal Investigator
 Gerard J. Montgomery
 (804) 977-0096
Business Contact
Phone: () -
Research Institution
N/A
Abstract
THE OBJECTIVE OF THIS EFFORT IS TO DEMONSTRATE A PROTOTYPE CONTROLLER FOR A HOWITZER USING A HYBRID NETWORK AND PRODUCTION RULE SYSTEM. THE RESULTING ADAPTIVE CONTROLLER WILL DEMONSTRATE THE FEASIBILITY OF DEVELOPING A REVOLUTIONARY CONTROLLER CAPABLE OF ADAPTING IN A DYNAMIC ENVIRONMENT, MONITORING AND DIAGNOSING BEHAVIOR, AND PREDICTING FUTURE FAULTS AND MAINTENANCE REQUIREMENTS. ABTECH HAS DEVELOPED AN INNOVATIVE NETWORK APPROACH, CALLED ABDUCTIVE NETWORKS, BASED ON MANY YEARS OF RESEARCH IN NEURAL NETWORKS AND ADVANCED STATISTICS. ABDUCTIVE NETWORKS ARE SIMPLY NETWORKS OF VERY POWERFUL NODE TYPES SUCH AS HIGH-ORDER POLYNOMIAL EQUATIONS. THIS PROPOSAL FIRST INTRODUCES AIM - A PRACTICAL SUPERVISED LEARNING TOOL FOR SYNTHESIZING ABDUCTIVE NETWORKS FROM DATABASES OF EXAMPLES. IT DISCUSSES HOW NETWORKS CAN BE APPLIED TO SUBSTANTIALLY INCREASE THE CONTROL AND DIAGNOSTIC CAPABILITIES OF SYSTEMS IN A PRACTICAL AND VERY COST-EFFECTIVE MANNER. IT THEN DETAILS A SPECIFIC PROJECT TO APPLY THE EXISTING AIM SOFTWARE AND A PRODUCTION RULE SYSTEM TO AN ARTILLERY SUBSYSTEM APPLICATION.

* information listed above is at the time of submission.

Agency Micro-sites

US Flag An Official Website of the United States Government