少しの工夫で大きな改善：シンプルな経路統合アルゴリズムによる経路品質の向上
A Little More, a Lot Better: Improving Path Quality by a Simple Path Merging Algorithm
複数の動作計画経路を動的計画法で統合し、経路長や滑らかさなどの品質を大幅に改善する簡便なアルゴリズムを提案した。12自由度までの問題で有効性を実証。
著者: Barak Raveh, Angela Enosh, Dan Halperin
分類: cs.RO, cs.AI
原文アブストラクト
Sampling-based motion planners are an effective means for generating collision-free motion paths. However, the quality of these motion paths (with respect to quality measures such as path length, clearance, smoothness or energy) is often notoriously low, especially in high-dimensional configuration spaces. We introduce a simple algorithm for merging an arbitrary number of input motion paths into a hybrid output path of superior quality, for a broad and general formulation of path quality. Our approach is based on the observation that the quality of certain sub-paths within each solution may be higher than the quality of the entire path. A dynamic-programming algorithm, which we recently developed for comparing and clustering multiple motion paths, reduces the running time of the merging algorithm significantly. We tested our algorithm in motion-planning problems with up to 12 degrees of freedom. We show that our algorithm is able to merge a handful of input paths produced by several different motion planners to produce output paths of much higher quality.
