Optimization-Based Maneuver Planning for a Tractor-Trailer Vehicle in a Curvy Tunnel: A Weak Reliance on Sampling and Search

This study is focused on the maneuver planning problem for a tractor-trailer vehicle in a curvy and tiny tunnel. Due to the curse of dimensionality, the prevalent sampling-and- search-based planners used to handle a rigid-body vehicle well become less efficient when the trailer number grows or when the tunnel narrows. This fact also has impacts on an optimization-based planner if it counts on a sampling-and-search-based initial guess to warm-start. We propose an optimization-based maneuver planner that weakly relies on the sampling and search, hoping to get rid of the curse of dimensionality and thus find optima rapidly. The proposed planner comprises three stages: stage I identifies the homotopy class via A(*) search in a 2D grid map; stage 2 recovers the kinematic feasibility with softened intermediate problems iteratively solved; stage 3 finds an optimum that strictly satisfies the nominal collision-avoidance constraints. Optimization-based planners are commonly known to run slowly, but this work shows that they have obvious advantages over the prevalent sampling-and-search-based planners when the solution space dimension is high and/or the constraints are harsh.

Automated summary

This study introduces an optimization-based maneuver planner that weakly relies on sampling and search, aiming to quickly find optimal solutions. The planner consists of three stages, utilizing A(*) search, iterative solving, and locating an optimum that strictly meets constraints. It demonstrates clear advantages over prevalent sampling-and-search-based planners in high-dimensional solution spaces and/or strict constraint scenarios.