arXiv subjects

Naira Hovakimyan

Publications and source records attributed to Naira Hovakimyan.

At least 37 records · Page 2Linked to original sources

arXiv2026

An Information-Theoretic Analysis of Continuous-Time Control and Filtering Limitations by the I-MMSE Relationships

While information theory has been introduced to characterize the fundamental limitations of control and filtering for a few decades, the existing information-theoretic methods are indirect and cumbersome for analyzing the limitations of continuous-time systems. To answer this challenge, we lift the information-theoretic analysis to continuous function spaces by the I-MMSE relationships. Continuous-time control and filtering systems are modeled into the additive Gaussian channels with and without feedback, and the total information rate is identified as a control and filtering trade-off metric and calculated from the estimation error of channel inputs. Fundamental constraints for this trade-off metric are first derived in a general setup and then used to capture the limitations of various control and filtering systems subject to linear and nonlinear plant models. For linear scenarios, we show that the total information rate quantifies the performance limits, such as the minimum entropy cost and the lowest achievable mean-square estimation error, in the time domain. For nonlinear systems, we provide a direct method to calculate and interpret the total information rate and its lower bound by the Stratonovich-Kushner equation.

cs.IT↗

arXiv2025

Distributionally Robust Imitation Learning: Layered Control Architecture for Certifiable Autonomy

Imitation learning (IL) enables autonomous behavior by learning from expert demonstrations. While more sample-efficient than comparative alternatives like reinforcement learning, IL is sensitive to compounding errors induced by distribution shifts. There are two significant sources of distribution shifts when using IL-based feedback laws on systems: distribution shifts caused by policy error and distribution shifts due to exogenous disturbances and endogenous model errors due to lack of learning. Our previously developed approaches, Taylor Series Imitation Learning (TaSIL) and $\mathcal{L}_1$ -Distributionally Robust Adaptive Control (\ellonedrac), address the challenge of distribution shifts in complementary ways. While TaSIL offers robustness against policy error-induced distribution shifts, \ellonedrac offers robustness against distribution shifts due to aleatoric and epistemic uncertainties. To enable certifiable IL for learned and/or uncertain dynamical systems, we formulate \textit{Distributionally Robust Imitation Policy (DRIP)} architecture, a Layered Control Architecture (LCA) that integrates TaSIL and~\ellonedrac. By judiciously designing individual layer-centric input and output requirements, we show how we can guarantee certificates for the entire control pipeline. Our solution paves the path for designing fully certifiable autonomy pipelines, by integrating learning-based components, such as perception, with certifiable model-based decision-making through the proposed LCA approach.

eess.SY↗

arXiv2025

Do Satellite Tasks Need Special Pretraining?

Foundation models have advanced machine learning across various modalities, including images. Recently multiple teams trained foundation models specialized for remote sensing applications. This line of research is motivated by the distinct characteristics of remote sensing imagery, specific applications and types of robustness useful for satellite image analysis. In this work we systematically challenge the idea that specific foundation models are more useful than general-purpose vision foundation models, at least in the small scale. First, we design a simple benchmark that measures generalization of remote sensing models towards images with lower resolution for two downstream tasks. Second, we train iBOT, a self-supervised vision encoder, on MillionAID, an ImageNet-scale satellite imagery dataset, with several modifications specific to remote sensing. We show that none of those pretrained models bring consistent improvements upon general-purpose baselines at the ViT-B scale.

cs.CV↗

arXiv2025

Real-Time Linear MPC for Quadrotors on SE(3): An Analytical Koopman-based Realization

This letter presents an analytical linear parameter-varying (LPV) representation of quadrotor dynamics utilizing Koopman theory, facilitating computationally efficient linear model predictive control (LMPC) for real-time trajectory tracking. By leveraging carefully designed Koopman observables, the proposed approach enables a compact lifted-space evolution that mitigates the curse of dimensionality while preserving the nonlinear characteristics of the system. Although model predictive control (MPC) is a powerful strategy for quadrotor control, it faces a trade-off between the high computational cost of nonlinear MPC (NMPC) and the reduced accuracy of LMPC. To address this gap, we introduce KQ-LMPC (Koopman Quasilinear LPV MPC), which leverages the Koopman-lifted LPV formulation to enforce constraints, ensure lower computational burden and real-time feasibility, and deliver tracking performance comparable to NMPC. Experimental validation confirms the effectiveness of the framework in reasonably agile flight. To the best of our knowledge, this is the first experimentally validated LMPC for quadrotors that employs analytically derived Koopman observables without requiring training data.

eess.SY↗

arXiv2025

A Simulation Evaluation Suite for Robust Adaptive Quadcopter Control

Robust adaptive control methods are essential for maintaining quadcopter performance under external disturbances and model uncertainties. However, fragmented evaluations across tasks, simulators, and implementations hinder systematic comparison of these methods. This paper introduces an easy-to-deploy, modular simulation testbed for quadcopter control, built on RotorPy, that enables evaluation under a wide range of disturbances such as wind, payload shifts, rotor faults, and control latency. The framework includes a library of representative adaptive and non-adaptive controllers and provides task-relevant metrics to assess tracking accuracy and robustness. The unified modular environment enables reproducible evaluation across control methods and eliminates redundant reimplementation of components such as disturbance models, trajectory generators, and analysis tools. We illustrate the testbed's versatility through examples spanning multiple disturbance scenarios and trajectory types, including automated stress testing, to demonstrate its utility for systematic analysis. Code is available at https://github.com/Dz298/AdaptiveQuadBench.

cs.RO↗

arXiv2025

Networked Control and Mean Field Problems Under Diagonal Dominance: Decentralized and Social Optimality

In this article, we employ an input-output approach to expand the study of cooperative multi-agent control and optimization problems characterized by mean-field interactions that admit decentralized and selfish solutions. The setting involves $n$ independent agents that interact solely through a shared cost function, which penalizes deviations of each agent from the group's average collective behavior. Building on our earlier results established for homogeneous agents, we extend the framework to nonidentical agents and show that, under a diagonal dominant interaction of the collective dynamics, with bounded local open-loop dynamics, the optimal controller for $H_\infty$ and $H_2$ norm minimization remains decentralized and selfish in the limit as the number of agents $n$ grows to infinity.

math.OC↗

arXiv2025

Flip Co-op: Cooperative Takeovers in Shared Autonomy

Shared autonomy requires principled mechanisms for allocating and transferring control between a human and an autonomous agent. Existing approaches often rely on blending control inputs between human and autonomous agent or switching rules, which lack theoretical guarantees. This paper develops a game-theoretic framework for modeling cooperative takeover in shared autonomy. We formulate the switching interaction as a dynamic game in which authority is embedded directly into the system dynamics, resulting in Nash equilibrium(NE)-based strategies rather than ad hoc switching rules. We establish the existence and characterization of NE in the space of pure takeover strategies under stochastic human intent. For the class of linear-quadratic systems, we derive closed-form recursions for takeover strategies and saddle-point value functions, providing analytical insight and efficient computation of cooperative takeover policies. We further introduce a bimatrix potential game reformulation to address scenarios where human and autonomy utilities are not perfectly aligned, yielding a unifying potential function that preserves tractability while capturing intent deviations. The framework is applied to a vehicle trajectory tracking problem, demonstrating how equilibrium takeover strategies adapt across straight and curved path segments. The results highlight the trade-off between human adaptability and autonomous efficiency and illustrate the practical benefits of grounding shared autonomy in cooperative game theory.

cs.HC↗

arXiv2025

Wasserstein Distributionally Robust Adaptive Covariance Steering

We present a methodology for predictable and safe covariance steering control of uncertain nonlinear stochastic processes. The systems under consideration are subject to general uncertainties, which include unbounded random disturbances (aleatoric uncertainties) and incomplete model knowledge (state-dependent epistemic uncertainties). These general uncertainties lead to temporally evolving state distributions that are entirely unknown, can have arbitrary shapes, and may diverge unquantifiably from expected behaviors, leading to unpredictable and unsafe behaviors. Our method relies on an $\mathcal{L}_1$-adaptive control architecture that ensures robust control of uncertain stochastic processes while providing Wasserstein metric certificates in the space of probability measures. We show how these distributional certificates can be incorporated into the high-level covariance control steering to guarantee safe control. Unlike existing distributionally robust planning and control methodologies, our approach avoids difficult-to-verify requirements like the availability of finite samples from the true underlying distribution or an a priori knowledge of time-varying ambiguity sets to which the state distributions are assumed to belong.

eess.SY↗

arXiv2025

$\mathcal{L}_1$-DRAC: Distributionally Robust Adaptive Control

Data-driven machine learning methodologies have attracted considerable attention for the control and estimation of dynamical systems. However, such implementations suffer from a lack of predictability and robustness. Thus, adoption of data-driven tools has been minimal for safety-aware applications despite their impressive empirical results. While classical tools like robust adaptive control can ensure predictable performance, their consolidation with data-driven methods remains a challenge and, when attempted, leads to conservative results. The difficulty of consolidation stems from the inherently different `spaces' that robust control and data-driven methods occupy. Data-driven methods suffer from the distribution-shift problem, which current robust adaptive controllers can only tackle if using over-simplified learning models and unverifiable assumptions. In this paper, we present $\mathcal{L}_1$ distributionally robust adaptive control ($\mathcal{L}_1$-DRAC): a control methodology for uncertain stochastic processes that guarantees robustness certificates in terms of uniform (finite-time) and maximal distributional deviation. We leverage the $\mathcal{L}_1$ adaptive control methodology to ensure the existence of Wasserstein ambiguity set around a nominal distribution, which is guaranteed to contain the true distribution. The uniform ambiguity set produces an ambiguity tube of distributions centered on the nominal temporally-varying nominal distribution. The designed controller generates the ambiguity tube in response to both epistemic (model uncertainties) and aleatoric (inherent randomness and disturbances) uncertainties.

eess.SY↗

arXiv2025

Observations Meet Actions: Learning Control-Sufficient Representations for Robust Policy Generalization

Capturing latent variations ("contexts") is key to deploying reinforcement-learning (RL) agents beyond their training regime. We recast context-based RL as a dual inference-control problem and formally characterize two properties and their hierarchy: observation sufficiency (preserving all predictive information) and control sufficiency (retaining decision-making relevant information). Exploiting this dichotomy, we derive a contextual evidence lower bound(ELBO)-style objective that cleanly separates representation learning from policy learning and optimizes it with Bottlenecked Contextual Policy Optimization (BCPO), an algorithm that places a variational information-bottleneck encoder in front of any off-policy policy learner. On standard continuous-control benchmarks with shifting physical parameters, BCPO matches or surpasses other baselines while using fewer samples and retaining performance far outside the training regime. The framework unifies theory, diagnostics, and practice for context-based RL.

cs.LG↗

arXiv2025

Bregman Centroid Guided Cross-Entropy Method

The Cross-Entropy Method (CEM) is a widely adopted trajectory optimizer in model-based reinforcement learning (MBRL), but its unimodal sampling strategy often leads to premature convergence in multimodal landscapes. In this work, we propose Bregman Centroid Guided CEM ($\mathcal{BC}$-EvoCEM), a lightweight enhancement to ensemble CEM that leverages $\textit{Bregman centroids}$ for principled information aggregation and diversity control. $\textbf{$\mathcal{BC}$-EvoCEM}$ computes a performance-weighted Bregman centroid across CEM workers and updates the least contributing ones by sampling within a trust region around the centroid. Leveraging the duality between Bregman divergences and exponential family distributions, we show that $\textbf{$\mathcal{BC}$-EvoCEM}$ integrates seamlessly into standard CEM pipelines with negligible overhead. Empirical results on synthetic benchmarks, a cluttered navigation task, and full MBRL pipelines demonstrate that $\textbf{$\mathcal{BC}$-EvoCEM}$ enhances both convergence and solution quality, providing a simple yet effective upgrade for CEM.

cs.LG↗

arXiv2025

Learning Tube-Certified Control using Robust Contraction Metrics

Control design for general nonlinear robotic systems with guaranteed stability and/or safety in the presence of model uncertainties is a challenging problem. Recent efforts attempt to learn a controller and a certificate (e.g., a Lyapunov function or a contraction metric) jointly using neural networks (NNs), in which model uncertainties are generally ignored during the learning process. In this paper, for nonlinear systems subject to bounded disturbances, we present a framework for jointly learning a robust nonlinear controller and a contraction metric using a novel disturbance rejection objective that certifies a tube bound using NNs for user-specified variables (e.g. control inputs). The learned controller aims to minimize the effect of disturbances on the actual trajectories of state and/or input variables from their nominal counterparts while providing certificate tubes around nominal trajectories that are guaranteed to contain actual trajectories in the presence of disturbances. Experimental results demonstrate that our framework can generate tighter (smaller) tubes and a controller that is computationally efficient to implement.

eess.SY↗

arXiv2025

DiffCoTune: Differentiable Co-Tuning for Cross-domain Robot Control

The deployment of robot controllers is hindered by modeling discrepancies due to necessary simplifications for computational tractability or inaccuracies in data-generating simulators. Such discrepancies typically require ad-hoc tuning to meet the desired performance, thereby ensuring successful transfer to a target domain. We propose a framework for automated, gradient-based tuning to enhance performance in the deployment domain by leveraging differentiable simulators. Our method collects rollouts in an iterative manner to co-tune the simulator and controller parameters, enabling systematic transfer within a few trials in the deployment domain. Specifically, we formulate multi-step objectives for tuning and employ alternating optimization to effectively adapt the controller to the deployment domain. The scalability of our framework is demonstrated by co-tuning model-based and learning-based controllers of arbitrary complexity for tasks ranging from low-dimensional cart-pole stabilization to high-dimensional quadruped and biped tracking, showing performance improvements across different deployment domains.

cs.RO↗

arXiv2025

Hessian Riemannian Flow For Multi-Population Wardrop Equilibrium

In this paper, we address the problem of optimizing flows on generalized graphs that feature multiple entry points and multiple populations, each with varying cost structures. We tackle this problem by considering the multi-population Wardrop equilibrium, defined through variational inequalities. We rigorously analyze the existence and uniqueness of the Wardrop equilibrium. Furthermore, we introduce an efficient numerical method to find the solution. In particular, we reformulate the equilibrium problem as a distributed optimization problem over subgraphs and introduce a novel Hessian Riemannian flow method, a Riemannian-manifold-projected Hessian flow, to efficiently compute a solution. Finally, we demonstrate the effectiveness of our approach through examples in urban traffic management, including routing for diverse vehicle types and strategies for minimizing emissions in congested environments.

eess.SY↗

arXiv2025

Task-Parameter Nexus: Task-Specific Parameter Learning for Model-Based Control

This paper presents the Task-Parameter Nexus (TPN), a learning-based approach for online determination of the (near-)optimal control parameters of model-based controllers (MBCs) for tracking tasks. In TPN, a deep neural network is introduced to predict the control parameters for any given tracking task at runtime, especially when optimal parameters for new tasks are not immediately available. To train this network, we constructed a trajectory bank with various speeds and curvatures that represent different motion characteristics. Then, for each trajectory in the bank, we auto-tune the optimal control parameters offline and use them as the corresponding ground truth. With this dataset, the TPN is trained by supervised learning. We evaluated the TPN on the quadrotor platform. In simulation experiments, it is shown that the TPN can predict near-optimal control parameters for a spectrum of tracking tasks, demonstrating its robust generalization capabilities to unseen tasks.

cs.RO↗

arXiv2025

An Information-Theoretic Analysis of Discrete-Time Control and Filtering Limitations by the I-MMSE Relationships

Fundamental limitations or performance trade-offs/limits are important properties and constraints of both control and filtering systems. Among various trade-off metrics, total information rate that characterizes the sensitivity trade-offs and time-averaged performance of control and filtering systems was conventionally studied by using the differential entropy rate and Kolmogorov-Bode formula. In this paper, by extending the famous I-MMSE (mutual information -- minimum mean-square error) relationships to the discrete-time additive white Gaussian channels with and without feedback, a new paradigm is introduced to estimate and analyze total information rate as a control and filtering trade-off metric. Under this framework, we explore the trade-off properties of total information rate for a variety of the discrete-time control and filtering systems, e.g., LTI, LTV, and nonlinear, and propose an alternative approach to investigate total information rate via optimal estimation.

eess.SY↗

arXiv2025

Coordinated Path Following of UAVs using Event-Triggered Communication over Networks with Digraph Topologies

This article presents a novel time-coordination algorithm based on event-triggered communication to ensure multiple UAVs progress along their desired paths in coordination with one another. In the proposed algorithm, a UAV transmits its progression information to its neighbor UAVs only when a decentralized trigger condition is satisfied. Consequently, it significantly reduces the volume of inter-vehicle communications required to achieve the goal compared with the existing algorithms based on continuous communication. With such intermittent communications, it is shown that a decentralized coordination controller guarantees exponential convergence of the coordination error to a neighborhood of zero. Furthermore, a lower bound on the difference between two consecutive event-triggered times is provided showing that the Zeno behavior is excluded with the proposed algorithm. Lastly, simulation results validate the efficacy of the proposed algorithm.

eess.SY↗

arXiv2025

$\mathcal{L}_{1}$ Adaptive Optimizer for Online Time-Varying Convex Optimization

We propose an adaptive method for online time-varying (TV) convex optimization, termed $\mathcal{L}_{1}$ adaptive optimization ($\mathcal{L}_{1}$-AO). TV optimizers utilize a prediction model to exploit the temporal structure of TV problems, which can be inaccurate in the online implementation. Inspired by $\mathcal{L}_{1}$ adaptive control, the proposed method augments an adaptive update law to estimate and compensate for the uncertainty from the prediction inaccuracies. The proposed method provides performance bounds of the error in the optimization variables and cost function, allowing efficient and reliable optimization for TV problems. Numerical simulation results demonstrate the effectiveness of the proposed method for online TV convex optimization.

math.OC↗