NASA 1998 SBIR Phase I


PROPOSAL NUMBER: 98-1 01.05-3155

PROJECT TITLE: Isolation-Enhanced Principle Component Analysis

TECHNICAL ABSTRACT (LIMIT 200 WORDS)

This Phase 1 SBIR proposal details an approach to equipment health monitoring which is able to diagnose a specific failure without an analytic model, and without data obtained for specific failure cases. It is quite common to add sensors to electromechanical equipment to detect and hopefully even diagnose failures. The basic idea of these approaches is to check whether the current measurement data taken from the Unit Under Test (UUT), on-line or off-line, are still in agreement with system nominal models which are representations of the relationships between inputs, disturbances, and outputs when the UUT is behaving normally, and under specific failure cases. If an inconsistency between a system nominal model and real-time measurements is detected, faulty components and sensors/actuators within the system need to be identified. Most systems in use today are model based, but the models are difficult to develop and are generally very imprecise. There are non-model based schemes but these require data from systems with specific failures to be able to diagnose failures. In practice, such failure data is very difficult to obtain. Our Isolation-enhanced PCA approach supports fault detection, and diagnosis of certain failures with no model, and without specific failure data. Only date for nominal operation is required.

POTENTIAL COMMERCIAL APPLICATIONS

Our commercialization goal is a set of software tools which allow detection and diagnosis locally or even across the Internet. Other tools we are developing provide support for remote repair actions, with video and audio provided across the Internet.

NAME AND ADDRESS OF PRINCIPAL INVESTIGATOR

Dr. Chiman Kwan
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