日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

週刊ニュースレター購読
arXiv:2310.11599

Hybrid Trajectory Optimization of Simple Skateboarding Tricks through Contact

Hybrid Trajectory Optimization of Simple Skateboarding Tricks through Contact

シェア:XThreadsFacebookLINEはてブBluesky

著者: Michael Burgess

分類: cs.RO, cs.SY, eess.SY

原文アブストラクト

Trajectories are optimized for a two-dimensional simplified skateboarding system to allow it to perform a fundamental skateboarding trick called an "ollie". A methodology for generating trick trajectories by controlling the position of a point-mass relative to a board is presented and demonstrated over a range of peak jump heights. A hybrid dynamics approach is taken to perform this optimization, with contact constraints applied along a sequence of discrete timesteps based on the board's position throughout designated sections of the trick. These constraints introduce explicit and implicit discontinuities between chosen sections of the trick sequence. The approach has been shown to be successful for a set of realistic system parameters.