arXiv subjects

Shankar A. Deka

Publications and source records attributed to Shankar A. Deka.

12 recordsLinked to original sources

arXiv2026

Estimating unknown dynamics and cost as a bilinear system with Koopman-based Inverse Optimal Control

In this work, we address the challenge of approximating unknown system dynamics and cost functions through a Koopman-based Inverse Optimal Control (IOC) framework. Using optimal trajectories, a modified Extended Dynamic Mode Decomposition with control (EDMDc) constructs a bilinear control system in lifted coordinates. Pontryagin's Maximum Principle (PMP) conditions are then derived, revealing structural similarities to the inverse Linear Quadratic Regulator (LQR) problem. This allows tractable cost recovery without resorting to nonlinear IOC formulations. The bilinear representation also inherits the analytical advantages of linear systems. Simulation and robotic experiments validate the approach, showing accurate estimation of both dynamics and costs, and illustrating its potential for general control and modeling applications.

eess.SY↗

arXiv2026

Data-driven Reachability Verification with Probabilistic Guarantees under Koopman Spectral Uncertainty

Providing rigorous reachability guarantees for unknown complex systems is a crucial and challenging task. In this paper, we present a novel data-driven framework that addresses this challenge by leveraging Koopman operator theory. Instead of operating in the state space, the proposed method encodes model uncertainty from finite data directly into Koopman spectral representation with quantifiable error bounds. Leveraging this spectral information, we systematically determine time intervals within which trajectories from the initial set are guaranteed, with a prescribed probability, to reach the target set. We finally demonstrate the efficacy of our framework in numerical examples.

eess.SY↗

arXiv2026

Differential Geometric Conditions for Koopman Linearizability of Control-Affine Systems

Koopman linearization opens many possibilities for control synthesis and analysis of nonlinear systems. Whether or not any given nonlinear control system admits a finite-dimensional Koopman representation remains a crucial question to address. A related problem is to categorize the class of all Koopman linearizable nonlinear control systems. In this work, we present differential geometric conditions on the drift and control vector fields of a control-affine nonlinear system, that must be necessarily satisfied for Koopman linear transformation to exist. The same conditions are also shown to be sufficient for (a slightly weaker notion of) Koopman linearizability on control-invariant manifolds. Further, these conditions, together with an additional condition, become necessary and sufficient for Koopman linearizability to a controllable linear system. Our examples illustrate the ease of checking these conditions, and also shed light on how Koopman linearizing transformation may not exist for a control-affine system even though one can linearize the autonomous part of the system via Koopman lifting.

math.OC↗

arXiv2026

Reach-Avoid Model Predictive Control with Guaranteed Recursive Feasibility via Input Constrained Backstepping

This letter proposes a novel sampled-data model predictive control framework for continuous control-affine nonlinear systems that provides rigorous reach-avoid and recursive feasibility guarantees under physical constraints. By propagating both input and output constraints through backstepping process, we present a constructive approach to synthesize a reach-avoid invariant set that complies with control input limits. Using this reach-avoid set as a terminal set, we prove that the proposed sampled-data MPC framework recursively admits feasible control inputs that safely steer the continuous system into the target set under fast sampling conditions. Numerical results demonstrate the efficacy of the proposed approach.

math.OC↗

arXiv2025

Time-to-reach Bounds for Verification of Dynamical Systems Using the Koopman Spectrum

In this work, we present a novel Koopman spectrum-based reachability verification method for nonlinear systems. Contrary to conventional methods that focus on characterizing all potential states of a dynamical system over a presupposed time span, our approach seeks to verify the reachability by assessing the non-emptiness of estimated time-to-reach intervals without engaging in the explicit computation of reachable set. Based on the spectral analysis of the Koopman operator, we reformulate the problem of verifying existence of reachable trajectories into the problem of determining feasible time-to-reach bounds required for system reachability. By solving linear programming (LP) problems, our algorithm can effectively estimate all potential time intervals during which a dynamical system can enter (and exit) target sets from given initial sets over an unbounded time horizon. Finally, we demonstrate our method in challenging settings, such as verifying the reachability between non-convex or even disconnected sets, as well as backward reachability and multiple entries into target sets. Additionally, we validate its applicability in addressing real-world challenges and scalability to high-dimensional systems through case studies in verifying the reachability of the cart-pole and multi-agent consensus systems.

eess.SY↗

arXiv2025

Robotic Trail Maker Platform for Rehabilitation in Neurological Conditions: Clinical Use Cases

Patients with neurological conditions require rehabilitation to restore their motor, visual, and cognitive abilities. To meet the shortage of therapists and reduce their workload, a robotic rehabilitation platform involving the clinical trail making test is proposed. Therapists can create custom trails for each patient and the patient can trace the trails using a robotic device. The platform can track the performance of the patient and use these data to provide dynamic assistance through the robot to the patient interface. Therefore, the proposed platform not only functions as an evaluation platform, but also trains the patient in recovery. The developed platform has been validated at a rehabilitation center, with therapists and patients operating the device. It was found that patients performed poorly while using the platform compared to healthy subjects and that the assistance provided also improved performance amongst patients. Statistical analysis demonstrated that the speed of the patients was significantly enhanced with the robotic assistance. Further, neural networks are trained to classify between patients and healthy subjects and to forecast their movements using the data collected.

cs.RO↗

arXiv2025

Backstepping Reach-avoid Controller Synthesis for Multi-input Multi-output Systems with Mixed Relative Degrees

Designing controllers with provable formal guarantees has become an urgent requirement for cyber-physical systems in safety-critical scenarios. Beyond addressing scalability in high-dimensional implementations, controller synthesis methodologies separating safety and reachability objectives may risk optimization infeasibility due to conflicting constraints, thereby significantly undermining their applicability in practical applications. In this paper, by leveraging feedback linearization and backstepping techniques, we present a novel framework for constructing provable reach-avoid formal certificates tailored to multi-input multi-output systems. Based on this, we developed a systematic synthesis approach for controllers with reach-avoid guarantees, which ensures that the outputs of the system eventually enter the predefined target set while staying within the required safe set. Finally, we demonstrate the effectiveness of our method through simulations.

eess.SY↗

arXiv2024

Extensions of the Path-integral formula for computation of Koopman eigenfunctions

Representing nonlinear dynamical systems using the Koopman Operator and its spectrum has distinct advantages in terms of linear interpretability of the model as well as in analysis and control synthesis through the use of well-studied techniques from linear systems theory. As such, efficient computation of Koopman eigenfunctions is of paramount importance towards enabling such Koopman-based constructions. To this end, several approaches have been proposed in literature, including data-driven, convex optimization, and Deep Learning-based methods. In our recent work, we proposed a novel approach based on path-integrals that allowed eigenfunction computations using a closed-form formula. In this paper, we present several important developments such as finite-time computations, relaxation of assumptions on the distribution of the principal Koopman eigenvalues, as well as extension towards saddle point systems, which greatly enhance the practical applicability of our method.

eess.SY↗

arXiv2023

Path-Integral Formula for Computing Koopman Eigenfunctions

The paper is about the computation of the principal spectrum of the Koopman operator (i.e., eigenvalues and eigenfunctions). The principal eigenfunctions of the Koopman operator are the ones with the corresponding eigenvalues equal to the eigenvalues of the linearization of the nonlinear system at an equilibrium point. The main contribution of this paper is to provide a novel approach for computing the principal eigenfunctions using a path-integral formula. Furthermore, we provide conditions based on the stability property of the dynamical system and the eigenvalues of the linearization towards computing the principal eigenfunction using the path-integral formula. Further, we provide a Deep Neural Network framework that utilizes our proposed path-integral approach for eigenfunction computation in high-dimension systems. Finally, we present simulation results for the computation of principal eigenfunction and demonstrate their application for determining the stable and unstable manifolds and constructing the Lyapunov function.

math.DS↗

arXiv2022

Koopman-Based Neural Lyapunov Functions for General Attractors

Koopman spectral theory has grown in the past decade as a powerful tool for dynamical systems analysis and control. In this paper, we show how recent data-driven techniques for estimating Koopman-Invariant subspaces with neural networks can be leveraged to extract Lyapunov certificates for the underlying system. In our work, we specifically focus on systems with a limit-cycle, beyond just an isolated equilibrium point, and use Koopman eigenfunctions to efficiently parameterize candidate Lyapunov functions to construct forward-invariant sets under some (unknown) attractor dynamics. Additionally, when the dynamics are polynomial and when neural networks are replaced by polynomials as a choice of function approximators in our approach, one can further leverage Sum-of-Squares programs and/or nonlinear programs to yield provably correct Lyapunov certificates. In such a polynomial case, our Koopman-based approach for constructing Lyapunov functions uses significantly fewer decision variables compared to directly formulating and solving a Sum-of-Squares optimization problem.

eess.SY↗

arXiv2021

Towards cyber-physical systems robust to communication delays: A differential game approach

Collaboration between interconnected cyber-physical systems is becoming increasingly pervasive. Time-delays in communication channels between such systems are known to induce catastrophic failure modes, like high frequency oscillations in robotic manipulators in bilateral teleoperation or string instability in platoons of autonomous vehicles. This paper considers nonlinear time-delay systems representing coupled robotic agents, and proposes controllers that are robust to time-varying communication delays. We introduce approximations that allow the delays to be considered as implicit control inputs themselves, and formulate the problem as a zero-sum differential game between the stabilizing controllers and the delays acting adversarially. The ensuing optimal control law is finally compared to known results from Lyapunov-Krasovskii based approaches via numerical experiments.

eess.SY↗

arXiv2020

Dynamically Computing Adversarial Perturbations for Recurrent Neural Networks

Convolutional and recurrent neural networks have been widely employed to achieve state-of-the-art performance on classification tasks. However, it has also been noted that these networks can be manipulated adversarially with relative ease, by carefully crafted additive perturbations to the input. Though several experimentally established prior works exist on crafting and defending against attacks, it is also desirable to have theoretical guarantees on the existence of adversarial examples and robustness margins of the network to such examples. We provide both in this paper. We focus specifically on recurrent architectures and draw inspiration from dynamical systems theory to naturally cast this as a control problem, allowing us to dynamically compute adversarial perturbations at each timestep of the input sequence, thus resembling a feedback controller. Illustrative examples are provided to supplement the theoretical discussions.

cs.LG↗