Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
To reject complex dynamically coupled disturbances without relying on restrictive fixed-decay priors, this paper develops a representation-learning framework that marries disturbance-accommodating control with modern last-layer adaptive methods. The approach leverages a hard expectation-maximization scheme with Kalman smoothing to learn latent disturbance dynamics constrained to be uniformly contractive, which are then tracked in real time using Bayesian filtering. Evaluated on a slippery ground vehicle burdened by coupled pendulum and liquid-sloshing dynamics as well as simulated Duffing oscillators, the method achieves superior tracking precision and disturbance prediction over standard LTI and fixed-decay adaptive baselines while ensuring provable exponential convergence.
Enforcing contractive dynamical structures onto learned latent disturbance representations delivers predictive disturbance accommodation without sacrificing the exponential stability guarantees essential for real-world adaptive control.
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, predictive DAC-style formulation. Combined with Bayesian filtering of the learned latent state, this representation yields a composite adaptive tracking controller with predictive capability and provable exponential convergence to a bounded neighborhood. We validate our approach experimentally on a slippery ground vehicle carrying a liquid-sloshing tank and a pendulum load, and we further assess its robustness on a system of coupled Duffing oscillators. Across both settings, the method achieves accurate disturbance prediction and improved overall tracking performance relative to fixed-decay representation-learning ablations, LTI disturbance-accommodating baselines, and model-based PD baselines.