Intelligent agents and hierarchical constraint-driven diagnostic units for a teleoperated fluid handling laboratory

Abstract

The purpose of this thesis is to study and develop intelligent agents for the forthcoming Space Station Freedom. Relevant intelligent capabilities, which are necessary in a semi-autonomous laboratory environment, are assumed to be built into a robot. An intelligent controller based on the DEVS formalism and the event-based approach is considered for an experiment. We shall discuss multiple model representations, where each model is tailored toward a specific purpose. Considering the necessity of diagnostic capabilities, we shall discuss the possibility of hierarchical diagnostic units for the Space Station. A high-level diagnostic unit is implemented on the basis of an artificial intelligence scheme and a hierarchy of diagnosers. This thesis also discusses the need for real-time diagnostic units and real-time data acquisition. We shall consider a constraint driven diagnostic unit which utilizes the time/cost (i.e., the actual associated cost or time in inquiring information necessary for a diagnosis process) criterion in an attempt to locate the cause(s) of failures.