Intelligent Embedded Diagnostic System for Future Avionic Systems

As modern electronic systems increase in complexity, it becomes more necessary to perform accurate health monitoring to ensure proper performance. Traditionally, crews of maintainers and technical experts have been sufficient to analyze faults throughobservation and electrical troubleshooting. However, this approach is unsuitable for modern avionics, in which thousands of electrical connections may exist between hundreds of avionic subsystems. As a result, significant effort has been expended todevelop automatic reasoning systems capable of sensing state parameters and making accurate decisions concerning the condition of the system. To reduce the complexity and cost associated with avionic system maintenance, Williams-Pyro proposes to continue development of its novel prognostics/diagnostics tool based on artificial neural network (ANN) technology. WPI has proven theeffectiveness of this concept within the Phase I effort and intends to further improve the Smart Embedded Avionic Monitoring System (SEAMS) to ensure operational suitability. Developed in conjunction with the University of Texas at Arlington, the SEAMSalgorithm combines the diagnostic capabilities of three different neural networks to produce a health monitoring analysis more accurate than any taken separately. The SEAMS system will provide an embedded means of detecting faults within avionic circuitry. If successful, this system will drastically reduce the cost associated with avionic maintenance - reducing the number of man-hours needed as well as the amountof test equipment. The initial target market for our Smart Embedded Avionic Monitoring System (SEAMS) is the U.S. Marine Corp's AH-1 Super Cobra Helicopter, the attack version of the UH-1