用向量量化压缩控制信号，让无人机在低算力下仍安全运行
Minimal Information Control Invariance via Vector Quantization
- 通过向量量化自动学习状态分区与有限控制码本
- 12维四旋翼模型中代码本规模缩小157倍仍保安全
- 适合关注安全控制与极简控制器的工程师
安全关键自主系统需在计算与感知资源受限下满足严格的状态约束，但基于学习的控制器通常远超安全运行所需复杂度。本文研究在采样数据控制下，使紧凑集合前向不变所需的最少不同控制信号数量，将其与信息论中的不变熵概念关联。提出一种向量量化自编码器，联合学习状态空间划分与有限控制码本，并开发基于Lipschitz可达集包络与平方和规划的迭代前向验证算法。在12维非线性四旋翼模型上，所学控制器相比均匀网格基线实现157倍的码本规模缩减，同时保持不变性，并实证刻画了安全运行兼容的最低感知分辨率。
原文摘要 · Abstract (English)
Safety-critical autonomous systems must satisfy hard state constraints under tight computational and sensing budgets, yet learning-based controllers are often far more complex than safe operation requires. To formalize this gap, we study how many distinct control signals are needed to render a compact set forward invariant under sampled-data control, connecting the question to the information-theoretic notion of invariance entropy. We propose a vector-quantized autoencoder that jointly learns a state-space partition and a finite control codebook, and develop an iterative forward certification algorithm that uses Lipschitz-based reachable-set enclosures and sum-of-squares programming. On a 12-dimensional nonlinear quadrotor model, the learned controller achieves a $157\times$ reduction in codebook size over a uniform grid baseline while preserving invariance, and we empirically characterize the minimum sensing resolution compatible with safe operation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。