ContactNets:滑らかな陰的表現で不連続な接触ダイナミクスを学習
ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations
接触による不連続な衝撃やスティクションを、物体間符号付き距離と接触フレームヤコビアンの滑らかな陰的表現として学習し、実世界60秒のデータで現実的な衝突・非貫入・静止摩擦を予測できる手法を提案した。
著者: Samuel Pfrommer, Mathew Halm, Michael Posa
分類: cs.RO, cs.LG
原文アブストラクト
Common methods for learning robot dynamics assume motion is continuous, causing unrealistic model predictions for systems undergoing discontinuous impact and stiction behavior. In this work, we resolve this conflict with a smooth, implicit encoding of the structure inherent to contact-induced discontinuities. Our method, ContactNets, learns parameterizations of inter-body signed distance and contact-frame Jacobians, a representation that is compatible with many simulation, control, and planning environments for robotics. We furthermore circumvent the need to differentiate through stiff or non-smooth dynamics with a novel loss function inspired by the principles of complementarity and maximum dissipation. Our method can predict realistic impact, non-penetration, and stiction when trained on 60 seconds of real-world data.
関連論文
- インスタンス非依存の幾何・接触ダイナミクス学習接触ダイナミクス
- 記号代数システムを活用した剛体システムの接触ダイナミクスシミュレーション接触ダイナミクス
- 物理構造化ニューラルネットワークによる接触ダイナミクスの学習接触ダイナミクス
