STATIC ATTITUDE ESTIMATION FROM LIGHT INTENSITY MEASUREMENTS: A HEURISTIC APPROACH

This paper explores the possibility of generating an attitude estimate for a space object from a set of static light intensity measurements using heuristic optimization techniques. The goal is to minimize the difference between a true observed intensity and an intensity calculated based on an estimated attitude along with knowledge of object geometry. Multiple measurements are required for observability, and multiple combinations of observers are considered. A particle swarm algorithm is implemented to identify the attitude that minimizes the intensity error. Several extensions to the standard particle swarm algorithm are considered, including a cluster-based global topology and cooling schedules for several heuristic parameters. Measurements are generated using the Phong anisotropic light reflection model. Once a good attitude estimate is obtained, it can be used to initialize an attitude filter. An unscented Kalman filter is implemented, and a range of angular rate errors are considered.