AI & ML
An {\alpha}-potential linear approximation for LQ differential games
arXiv cs.SY0 views
arXiv:2611.25149v1 Announce Type: new Abstract: Feedback Nash equilibria of infinite-horizon linear-quadratic (LQ) differential games are characterized by coupled algebraic Riccati equations (cAREs), whose numerical solution generally relies on iterative algorithms and can be highly sensitive to their initialization. Motivated by the $\alpha$-potential-game framework, this letter proposes an $\alpha$-potential linear approximation for constructing a systematic, game-dependent initialization. For an N-player LQ game, a reference potential game is constructed. Its equilibrium requires solving only one standard algebraic Riccati equation. The original cAREs are subsequently linearized around this reference solution, yielding a coupled linear system whose solution provides player-specific corrections and hence an approximate feedback Nash equilibrium. The resulting approximation is used to initialize three standard cARE solvers. Numerical experiments on $10^4$ randomly generated two-player games show that the proposed initialization reduces the mean number of iterations. Further experiments for $N=2,\ldots,13$ show that the reduction in iteration count persists as the number of players increases.
Source:https://arxiv.org/abs/2611.25149PreviousAsymmetric Transistor Aging Under Thermal GradientsNextReliability-based Sizing of Energy Storage for Large Load Integration
Related posts
Reading project health off a download page — faster than star count
0dev.to
Best time trackers for developers, for code and for AI
0dev.to

Serving Qwen3.5-0.8B with a 0.016% weight edit: measure the loops, pin the sampling on the server
0dev.to
Design Tokens for AI: Extract a Website's Real Colours, Fonts, Shadows and Icons So an LLM Builds On Brand
0dev.to
Trying to get our Arms around AI efficiency
0Stack Overflow Blog
Why an Embedded Game Fails on Some Phones, and How to Debug It
0dev.to