R W. Johnson and S. Jayaram
High Fidelity Dynamic Mo deling (HFDM), real-time, Control Station (SGCS) Mechanical, Materials and Aerospace Department, P.O. Box 162450, Orlando, FL 32816-2450 USA; e-mail: {rjohnson, sjayaram} @p e
A new real-time detection/diagnosis methodology for an Automated Ground Health Monitoring System (AGHMS) is applied at a satellite ground control station. The technological innovations in this research are fo cused on the identification of abnormal transient resp onse profiles from a satellite 6-DOF attitude control platform. The identification will b e made by comparing, in real time, the filtered (Kalman) measurements to a synchronized mo del reference system. The methodology used to accomplish this task will b e software intensive and p erfectly compatible with the op en physical architecture of existing monitoring devices and their automated control system mechanisms. The innovations demonstrated will b e (1) a real-time Extended Kalman filter to eliminate the measurement noise; (2) the formulation and use of a "dynamic threshold" detection system to identify abnormal "state estimates" as well as "covariance estimates"; and (3) the generation of an intelligent fault-mo de file with corrective control commands to stabilize mild detected faults. The ob jective of this article is to provide these technical enhancements by handling and evaluating test data differently. An example is included to demonstrate these technical innovations. The AGHMS metho dology will demonstrate real-time signal detection using an Extended Kalman filter (EKF) to obtain the b est estimate of measurements (i.e., the dynamic parameters of a system); to obtain precise knowledge of the attitude of a satellite or a spacecraft that has an onb oard magnetometer for attitude measurements; to study and analyze the state co-variance, as well as the error co-variance of the system; and to enhance the processing capability by monitoring systems in real time, perform systems detection/diagnosis, and actively control the environment/process based on these onb oard sensor readings.1. Intro duction This article focuses on the identification/demonstration of critical technology innovations that will be applied to various applications, such as detection/diagnosis of automated mechanisms for health monitoring, Real-time data analysis, and the control of various systems and their subsystems. This new innovation using a High Fidelity, Dynamic Model-based Simulation (HFDMS) approach will be used to implement a real-time monitoring, test and evaluation methodology. The unique element of this process control technique is the use of high-fidelity, computer generated dynamic models to replicate the behavior of actual systems under test (SUT). It will provide a dynamic simulation capability that becomes the reference truth model, from which comparisons are made with the actual raw/unconditioned data from test elements. The insertion of this new concept of health monitoring into existing automated monitoring and control systems will provide a real-time, intelligent command and control system that has the capability to monitor and observe transient behaviour and relate that behaviour to the dynamic parameters of the systems being tested. Current test capability cannot measure the dynamic behaviour of SUT in real time. Abnormal dynamic responses are indicators of an out-of-tolerance performance of the SUT; they can be a predictor of impending failures in those systems. This feature adds a new dimension to existing test control mechanizations; it will greatly enhance the accuracy of the "system state," which in turn will increase the reliability of the test and evaluation process over those currently in use. This processing technique triggers the real-time detection of abnormal data flow conditions and will inititate automatic identification of the specific "state" causing the fault condition. This attribute will speed up diagnostic analysis to seconds rather than minutes/hours, thus reducing significantly fault detection, diagnosis, and any indicated control corrections. The mechanization of Satellite Health Monitoring (SHM) will be initially incorporated into a Satellite Ground Control Station (SGCS). A High-Fidelity Dynamic Modeling (HFDM) technique will be used to implement a realtime methodology for a spacecraft/satellite. This process 27
Important Links:
Go Back
