NASA 1998 SBIR Phase I


PROPOSAL NUMBER: 98-1 01.04-3155A

PROJECT TITLE: Model-Based Fault Detection and Identification

TECHNICAL ABSTRACT (LIMIT 200 WORDS)

The innovation herein detailed is a technique for designing model-based fault detection and isolation (FDI) for Guidance, Navigation, and Control (GNC) at the system level. Most model-based FDI approaches use the system dynamic models directly. The problems with these approaches are that the system dynamic models are very complicated and the system topology may not reflect the actual flow of failure information. Our FDI approach is based on "Dependency Model" (DM), which describes the high level information flow during the occurrence of failures or faults. Therefore, DM is more suitable for FDI than the system dynamic model. The most difficult task in DM-based FDI is the generation of the DM. For example, it usually takes several men/months to generate a DM for a regular avionics system. In this Phase I work, we propose to develop a technique to automatically derive a DM from a GNC system dynamic model described in a system level simulator such as MATLAB and SIMULINK. When integrated with testability analysis tool developed by IAI from previous works, we can have an automatic tool for designing FDI for GNC. The result of this research can reduce the time and cost of FDI development significantly.

POTENTIAL COMMERCIAL APPLICATIONS

This FDI design tool will find a big commercial market in the airplane industry and government agencies including NASA, Air Force, Army, and Navy. We can package it as a software package to be used by clients, or we can use it to provide FDI design services for customers. The proposed design tool can be extended to other parts of spacecraft systems, such as propulsion system, environmental control and life support system, hydraulic system, electrical power system, attitude control system, etc. In addition, the same technology can be easily applied to ground and ocean vehicles. Therefore, the proposed design tool has a huge commercialization potential.

NAME AND ADDRESS OF PRINCIPAL INVESTIGATOR

Dr. Chujen Lin
Intelligent Automation, Inc.
2 Research Place, Suite 202
Rockville , MD 20850

NAME AND ADDRESS OF OFFEROR

Intelligent Automation, Inc.
2 Research Place, Suite 202
Rockville , MD 20850