Safe Motion Planning Against Multimodal Distributions Based on a Scenario Approach

We present the design of a motion planning algorithm that ensures safety for an autonomous vehicle. In particular, we consider multimodal distribution over uncertainties; example, uncertain predictions future trajectories surrounding vehicles reflect discrete decisions, such as turning or going straight at intersections. develop computationally efficient, scenario-based approach solves problem with high confidence given quantifiable number samples from distribution. Our is based on two preprocessing steps, which 1) separate into distinct clusters and 2) compute bounding polytope each cluster. Then, rewrite approximately mixed-integer using polytopes. demonstrate via simulation nuScenes dataset our probability in presence uncertainties, more efficient less conservative than conventional scenario approach.