This work presents an adaptive, information-based approach to dynamic sensor network management to track multiple maneuvering satellites with a diversely populated Space Object Surveillance and Identification network. Previous sensor tasking strategies, which rely on traditional orbit determination methods, will often fail when attempting to track maneuvering targets. The proposed method integrates a Multiple-Model Adaptive Un-scented Kalman Filter, a Largest Lyapunov Exponent approximation metric, and either Fisher or Shannon Information Gain to prioritize target spacecraft and task sensors. The algorithm must manage a globally distributed network of ground and space-based radar and electro-optical sensors. The sensor network must monitor a constellation of potentially maneuvering spacecraft that span all inclinations and altitudes up to geosynchronous orbit. The results show that Multiple-Model Adaptive Estimation coupled with information gain metrics can effectively task a diverse network of sensors to track multiple maneuvering spacecraft, while simultaneously monitoring a large number of non-maneuvering objects.
展开▼