[论文] Statistical Learning of Contractive Dynamical Representations for Comp...

论文概要

研究领域: ML 作者: Min Kim, José Leonardo Brenes, Fred Hadaegh, Soon-Jo Chung 发布时间: 2026-09-28 arXiv: 2609.35758

中文摘要

我们提出了一个用于动态耦合干扰下复合自适应跟踪控制的表示学习框架，将经典干扰调节控制（DAC）与最新的末层自适应干扰抑制方法联系起来。具体而言，我们引入一种统计学上严谨的硬期望最大化（hard-EM）过程，在硬 E 步中使用 Kalman 平滑器，以识别潜在演化一致收缩的动态干扰表示。学习到的表示从测量的被控对象特征和控制输入演化潜在干扰激励状态，并将该状态解码为作用于标称被控对象的时变干扰，从而将先前的固定衰减末层自适应方法扩展为学习到的、预测性的 DAC 风格公式。结合对学习到的潜在状态的贝叶斯滤波，该表示产生具有预测能力和可证明指数收敛到有界邻域的复合自适应跟踪控制器。我们在携带液体晃动罐和钟摆负载的湿滑地面车辆上实验验证，并在耦合 Duffing 振子系统上评估鲁棒性。在两种场景中，该方法均实现了准确的干扰预测和优于固定衰减表示学习消融、LTI 干扰调节基线和基于模型的 PD 基线的整体跟踪性能。

原文摘要

We present a representation-learning framework for composite adaptive tracking control under dynamically coupled disturbances. The framework connects classical disturbance-accommodating control (DAC) to recent last-layer adaptive disturbance-rejection methods. Specifically, we introduce a statistically principled hard expectation-maximization (hard-EM) procedure, with a Kalman smoother in the hard E-step, to identify dynamical representations of disturbance whose latent evolution is uniformly contractive. The learned representation evolves a latent disturbance-excitation state from measured plant features and control inputs and decodes that state into the time-varying disturbance acting on the nominal plant, thereby extending prior "fixed-decay" last-layer adaptive methods to a learned, pr...

#论文 #arXiv #ML #小凯