An Agent-based Approach to Health Monitoring Systems Applied to the ISS Power System

Abstract

We report on the design and development of an agent-based approach for health monitoring systems. A health monitoring system typically involves multiple sensors whose readings provide information as to the status of equipment components. Abnormal sensor readings can indicate malfunctions, either in a particular piece of equipment or in another component on which it depends or with which it interacts. Health monitoring systems operate in near real time. Large or complex systems that involve numerous sensors are data intensive. Various algorithms or data mining approaches support the identification of anomalous system operation, as well as classification of probable cause; i.e., what has failed. As physical systems become more complex and/ or more autonomous, health monitoring systems become more critical. Our work has focused on the design and development of an agent-based health monitoring system that serves as a conceptual prototype for autonomous health monitoring systems for support of space exploration applications. We utilized the electrical power system (EPS) of the International Space Station (ISS) as the physical system of interest. Pertinent data used in our health monitoring system is provided by results from a high fidelity simulation of the ISS Power System. Our agent-based health monitoring (ABHM) system design provides for flexibility; multiple families of assessment algorithms can be applied to the sensor data in parallel.