This work describes trajectory planning for autonomous driving. It treats continuous methods, which converge to a local optimum, as well as combinatorial methods, which are globally optimal. It is shown by analysis and experimentally, that both approaches are suitable for autonomous driving, but cover different driving situations.
