It has long been recognized that automation will play an important role in the Space Station program. Many benefits accrue for the utilization of computer-based automation technology in terms of the management, operation, monitoring, and control of complex Space Station subsystems. On the Space Station, the electric power system plays a key role in coordinated operation with the other on-board subsystems. It supplies the primary resource (electric power) upon which all the other subsystems and experiments rely. Because of the unique and complex role that the electric power system will play, there is a great potential for applying cooperative knowledge-based system technology to increase reliability and lower operating costs.The short-term objective of the work described in this article is to develop ground-and knowledge-based systems integrated with actual electric power system breadboards and test beds to demonstrate the viability of such advanced automation approaches for spacecraft on-board and ground-support applications. Initially, such systems would be primarily employed in advisory capacities. As confidence is gained in their operation, these systems would evolve to enable closedloop control. The long-term objective is to develop such intelligent knowledge-based systems for actual on-board autonomous operation of the spacecraft electric power system.In the Electrical Power Branch at Marshall Space Flight Center (MSFC), attention has been focused on comprehensive fault management, including dynamic payload rescheduling activities. Comprehensive fault management includes identifying anomalies, diagnosing actual faults (hard and soft), isolation of faults, recommending corrective actions for fault recovery, and autonomous implementation of fault recovery actions. Dynamic payload rescheduling is necessary in order to operate a closed-loop autonomous electric power system of significant size and complexity.When a fault develops (or begins to develop as an incipient failure), the system should be able to identify what is happening, to locate the problem source, to recommend actions to be taken, to implement those actions autonomously, to evaluate the payload schedule for perturbations, and to reschedule payloads consistent with the current electric power system configuration following recovery actions taken by the system. The approach taken within the Information and Electronic Systems Laboratory at MSFC has been to start with well-defined, limited electric power system applications as stand-alone systems. The next step has been to integrate such applications with actual breadboards that are representative of flight systems. The next activity is concemed with focusing on integrating these knowledge-based systems with conventional automation software and hardware. Current efforts emphasize adding robustness to the interactions among the knowledge-based systems as well as adding intermediate levels of autonomy for the power system management. The Electrical Power Branch at MSFC has been involved with the developme...
