Massively Parallel Optimization of Target Sequences for Multiple-Rendezvous Low-Thrust Missions on GPUs

In this work a massively parallel method for identification of optimal sequences of targets in multiple-rendezvous low-thrust missions using GPU processors is presented. Given a list of possible targets, an exhaustive search of sequences compatible with mission requirements is performed. To estimate feasibility of each transfer, a heuristic model based on Lambert transfers is evaluated in parallel. The resulting sequences are ranked by user-specified criteria such as length or fuel consumption. The algorithm has been used to compute asteroid sequences for GTOC7. The efficiency of the GPU implementation is demonstrated by comparing it with a traditional CPU based branch and bound method.