AI & ML
Lie Group Structure of Six-Degree-of-Freedom Rigid-Body Uncertainty Representation and Propagation
arXiv eess.SY0 views
arXiv:2611.19000v2 Announce Type: replace Abstract: For six-degree-of-freedom rigid-body motion, the choice of Lie-group structure affects both how well the group-theoretic mean and covariance represent the state distribution and how accurately these moments are propagated. A quaternion-based, left-trivialized semidirect-product formulation of the velocity phase space is developed for the actuated body-frame dynamics, with closed-form group operations and Jacobians that avoid matrix-group embeddings. We derive the controlled stochastic dynamics and establish that the direct- and semidirect-product formulations describe the same physical process under the specified noise model. Since the semidirect exponential coordinates account for frame changes between body twists at different poses, they capture pose--twist coupling more naturally. A lunar powered-descent experiment assesses low-order moment representations of uncertainty and propagation performance under the proposed group structure using two distinct algorithms. Compared with direct-product formulations and a linearized baseline, the proposed structure improves the representation of the banana-shaped body-velocity distribution using group-theoretic moments, yields more consistent propagated moments, and produces distribution approximations that agree more closely with the Monte Carlo reference. In this experiment, changing the group structure accounts for most of the observed improvement, highlighting the practical value of the proposed semidirect-product formulation.
Source:https://arxiv.org/abs/2611.19000PreviousEvery minute counts: Wind farm power control during imbalance settlement using distributional soft actor-criticNextSocial Welfare and Equilibrium in Distributed Microgrid Prosumer Coordination Under a Mixed Competitive-Cooperative Setting
Related posts
How to detect user frustration in chatbot conversations
0dev.to
Mình Đã Biến Gemini Thành Trợ Lý Cá Nhân Như Thế Nào (Lịch, Drive, Gmail và Những Lần Suýt Toang)
0dev.to
Im looking fr any free ai models / tools are there?, leaVE UR Lights💡
0dev.to
I checked 10 SEO agency websites for AI-search readiness. The real bugs were the boring ones.
0dev.to

How much GPU memory does an LLM need? The whole sum, worked out on real cards
0dev.to

Routing Upscaling Jobs by Output Size, Not Input Size
0dev.to