DYNAMIC SENSOR STEERING FOR TARGET SEARCH FOR SPACE SITUATIONAL AWARENESS

This paper proposes a set of particle clustering-based sensor steering algorithms to search for a target among a large set of candidate orbits in the context of space situational awareness (SSA). The challenge of such an initial target problem is that the uncertainty region of the target is much larger than the sensor's FOV and that the sensor is not perfect. Existing initial target search methods use particle-based methods to represent the uncertainties and evaluate each particles to determine the sensor pointing at every time step. The proposed algorithms improve the traditional particle-based algorithms by clustering the particles so that we only need to evaluate the cluster centroids to determine the sensor pointing instead of evaluating all particles. Simulations show that this approach can save the computational cost of particle-based methods significantly while achieving a similar initial target search performance compared to the traditional methods. The proposed methods are expected to improve the real-time performance of initial target search.