SATELLITE COLLISION ANALYSIS USING GENETIC ALGROITHMS AS A FILTER

摘要

When doing collision analysis, it is important to compare close approaches between all active satellites in space with all tracked objects in space. All possible combination of object pairs must be analyzed. An algorithm for finding object pairs that have a low chance of collision has been developed. To allow for quick computation so that all collision possibilities do not need to be analyzed using a high-order propagator, a low-order propagator that is easily parallelizable incorporates a genetic algorithm to find closest approach. The result is a parallel algorithm that filters out low probability collision pairs thus reducing the computation time necessary to evaluate the overall close-approach risk for all of the object pairs.