Collision avoidance in satellite clusters

Abstract

A collision avoidance methodology for satellite clusters is developed and simulated. A set membership algorithm is used to monitor and predict guaranteed position uncertainty ellipsoids over time. Distances between ellipsoids can be calculated on-line in order to evaluate when and how fast a collision may occur. The algorithm assumes guaranteed knowledge of an initial state (relative position and velocity) error, sensor error, and forcing function. The forcing function is variable depending on orbital parameters and faults. A probability of collision is evaluated using the resulting ellipsoid information. Results show that satellite clusters with a close proximity will require a back-up system that is accurate and gives measurement updates at least every few minutes.