AI & ML
Minimum-Time Escape from the Conic Field of View of a Reeds-Shepp Pursuer: Classification and Game Value for a Slower or Equally Fast Evader
arXiv cs.RO0 views
arXiv:2610.27629v1 Announce Type: new Abstract: We study minimum-time escape of an omnidirectional evader from the forward conic field of view of a Reeds-Shepp pursuer with minimum turning radius $r$. For speed ratio $\nu$ and cone half-angle $0<\alpha<\pi/2$, we prove a global escape/holding classification: straight reversal holds the whole cone for $\nu\le\sin\alpha$, escape is confined to a barrier-bounded dart for $\sin\alpha<\nu\le1$, and escape is possible from every state for $\nu>1$. For the slower or equally fast evader, $\sin\alpha<\nu\le1$, we construct the solution with Isaacs' characteristics and verify it in feedback form, using a Filippov pursuer relay and a one-sided evader branch. The finite escape time is the first root of one scalar equation, and the holding region is bounded by the nonusable parts of the rays and a barrier that closes with a convex corner on the front axis. Sample-and-hold counter-controls show that this escape time is the value of the game in nonanticipative strategies at every state except the corner. There the two strategy values differ: the pursuer's is infinite, whereas an evader that chooses which barrier branch to cross from the initial germ of the pursuer's control escapes before the barrier time $\tau_I$, which is the evader's exact strategy value. An independent level-set solver, refined on fixed comparison points, agrees with the closed-form escape time and reproduces the holding region at the tested parameters.
Source:https://arxiv.org/abs/2610.27629PreviousA Head Camera Is (Almost) All You Need: An Electronics-Free UMI Glove Labelled and Observed by One Moving CameraNextPLaCA: Learning Patch-Level Future Latent Change for Action Generation
Related posts
Most voted crypto projects on Coinvote, Oct 26 to Nov 1, 2026
0dev.to Data
Bluesky This Week — trending topics & AI brand mentions (2026-11-02)
0dev.to Data
Gemini 4 Carbon e DeepSeek V4.1 Flash Cyber: a Guerra Cibernética Autônoma e o Topo do Artificial Analysis
0dev.to Machine Learning
What can an AI agent do for you every day?
0dev.to Machine Learning
Mon modèle de paris semblait excellent sur 3 671 prédictions. Il n'a jamais battu le bookmaker.
0dev.to Machine Learning
HR manager LinkedIn About section examples you can adapt
0dev.to Career