Project Title:

A Hybrid Analytical-Intelligent Approach to Fault-Tolerant Control of Large Space

09.01-9995
900800
A Hybrid Analytical-Intelligent Approach to Fault-Tolerant Control of Large Space

Abstract:

Structures
This project combines hybrid control-theoretic techniques and artificial-intelligence
for the design of fault-tolerant control strategies for large space structures. Available
AI-based or purely control-theoretic techniques for fault-tolerant control offer
no unified methodology to integrate the diverse issues of system modeling, fault
detection and isolation, fault propagation, system restructuring, and controller
reconfiguration. This project capitalizes upon structural features of the system.
The fault-tolerant design procedure entails the combination of signal redundancy
and fuzzy logic. When a fault is detected, causal or qualitative reasoning is used
to model its propagation to adjacent subsystems. The system is restructured next
by isolating faulty components. Finally, a design method is suggested leading to
a structural control law that reconfigures the original system controller in order
to meet the primary objective of guaranteed stability while the system is operating
in a degraded mode. Algorithmic developments are to be demonstrated using the Thermal
Control System of the space station's common module as the test bed.
This project will introduce innovative methods to the fault-tolerant control of space
vehicles and other complex processes. Software packages and controller-sensor-actuator
configurations would be provided for a variety of applications.
fault-tolerant control, large-scale systems, reconfigurable control