Papers

When electric vehicles (EVs) are integrated into standalone DC microgrids (DCMGs), stability issues arise due to their constant power load (CPL) behavior, which provides negative incremental impedance (NII). In addition, the microgrids suffer from an inherent low-inertia problem. Therefore, this study presents a composite controller incorporating a global integral terminal sliding mode controller with a backstepping controller. A virtual capacitor is employed to mitigate the low-inertia issue and strengthen the DC-bus response. An improved fractional power-based reaching law decreases chattering and accelerates convergence. Exact feedback linearization converts the nonlinear boost converter model into Brunovsky's canonical form, mitigating NII effects and non-minimum phase issues. The entire system stability is verified using Lyapunov control theory. Simulation outcomes confirm superior performance, with 34.4-53.3% reduction in overshoot, 52.9-74.9% in undershoot, and 12-47.4% in settling time compared to the existing controller.

Controlling high-dimensional stochastic systems, critical in robotics, autonomous vehicles, and hyperchaotic systems, faces the curse of dimensionality, lacks temporal abstraction, and often fails to ensure stochastic stability. To overcome these limitations, this study introduces the Multi-Timescale Lyapunov-Constrained Hierarchical Reinforcement Learning (MTLHRL) framework. MTLHRL integrates a hierarchical policy within a semi-Markov Decision Process (SMDP), featuring a high-level policy for strategic planning and a low-level policy for reactive control, which effectively manages complex, multi-timescale decision-making and reduces dimensionality overhead. Stability is rigorously enforced using a neural Lyapunov function optimized via Lagrangian relaxation and multi-timescale actor-critic updates, ensuring mean-square boundedness or asymptotic stability in the face of stochastic dynamics. The framework promotes efficient and reliable learning through trust-region constraints and decoupled optimization. Extensive simulations on an 8D hyperchaotic system and a 5-DOF robotic manipulator demonstrate MTLHRL's empirical superiority. It significantly outperforms baseline methods in both stability and performance, recording the lowest error indices (e.g., Integral Absolute Error (IAE): 3.912 in hyperchaotic control and IAE: 1.623 in robotics), achieving faster convergence, and exhibiting superior disturbance rejection. MTLHRL offers a theoretically grounded and practically viable solution for robust control of complex stochastic systems.

Commercial off-the-shelf (COTS) Global Navigation Satellite System (GNSS) receivers face significant limitations under high-dynamic conditions, particularly in high-acceleration environments such as those experienced by launch vehicles. These performance degradations, often observed as discontinuities in the navigation solution, arise from the inability of traditional tracking loop bandwidths to cope with rapid variations in synchronization parameters. Software-Defined Radio (SDR) receivers overcome these constraints by enabling flexible reconfiguration of tracking loops; however, manual tuning involves a complex, multidimensional search and seldom ensures optimal performance. This work introduces a genetic algorithm-based optimization framework that autonomously explores the receiver configuration space to determine optimal loop parameters for phase, frequency, and delay tracking. The approach is validated within an SDR environment using realistically simulated GPS L1 signals for three representative dynamic regimes -guided rocket flight, Low Earth Orbit (LEO) satellite, and static receiver-processed with the open-source GNSS-SDR architecture. Results demonstrate that evolutionary optimization enables SDR receivers to maintain robust and accurate Position, Velocity, and Time (PVT) solutions across diverse dynamic conditions. The optimized configurations yielded maximum position and velocity errors of approximately 6 m and 0.08 m/s for the static case, 12 m and 2.5 m/s for the rocket case, and 5 m and 0.2 m/s for the LEO case.

Our infrastructure systems enable our well-being by allowing us to move, store, and transform materials and information given considerable social and environmental variation. Critically, this ability is shaped by the degree to which society invests in infrastructure, a fundamentally political question in large public systems. There, infrastructure providers are distinguished from users through political processes, such as elections, and there is considerable heterogeneity among users. Previous political economic models have not taken into account (i) dynamic infrastructures, (ii) dynamic user preferences, and (iii) alternatives to rational actor theory. Meanwhile, engineering often neglects politics. We address these gaps with a general dynamic model of shared infrastructure systems that incorporates theories from political economy, social-ecological systems, and political psychology. We use the model to develop propositions on how multiple characteristics of the political process impact the robustness of shared infrastructure systems to capacity shocks and unequal opportunity for private infrastructure investment. Under user fees, inequality decreases robustness, but taxing private infrastructure use can increase robustness if non-elites have equal political influence. Election cycle periods have a nonlinear effect where increasing them increases robustness up to a point but decreases robustness beyond that point. Further, there is a negative relationship between the ideological sensitivity of candidates and robustness. Overall, the biases of voters and candidates (whether they favor tax increases or decreases) mediate these political-economic effects on robustness because biases may or may not match the reality of system needs (whether system recovery requires tax increases).

This work presents a purely data-driven, wavelet-based framework for modal identification and reduced-order modeling of mechanical systems with assumed linear dynamics characterized by closely spaced modes with classical or non-classical damping distribution. Traditional Fourier-based methods often fail to reliably identify closely spaced modes or accurately capture modal interactions and complexities. To address these limitations, we propose a methodology leveraging the enhanced time -frequency resolution capabilities of the continuous wavelet transform (CWT). By selecting appropriate harmonic regions within the wavelet spectra, we effectively isolate modes, and then invert them back in the temporal domain by applying the inverse CWT (ICWT). In this way we reconstruct the corresponding modal dynamics in the time domain. Using the Hilbert transform, instantaneous phases are extracted for each identified mode, enabling the introduction of a complexified modal matrix which robustly characterizes the system's modal properties, even under challenging perturbations such as noise and uncertainties due to modal interference and unmodeled effects. The identified modal parameters are utilized to reconstruct the frequency response functions (FRFs) of the system and to develop a reduced-order model (ROM) that captures accurately the system's dominant dynamical behavior valid in a specified frequency range.. Validation of the methodology is conducted both with a numerical non-classical damping and an experimental testbed representing a model of an airplane structure. Results demonstrate the effectiveness of the proposed approach in resolving intricate modal interactions and accurately reproducing the dynamic response of complex structural systems.

Image-to-image translation models have achieved notable success in converting images across visual domains and are increasingly used for medical tasks such as predicting post-operative outcomes and modeling disease progression. However, most existing methods primarily aim to match the target distribution and often neglect spatial correspondences between the source and translated images. This limitation can lead to structural inconsistencies and hallucinations, undermining the reliability and interpretability of the predictions. These challenges are accentuated in clinical applications by the stringent requirement for anatomical accuracy. In this work, we present TraceTrans, a novel deformable image translation model designed for post-operative prediction that generates images aligned with the target distribution while explicitly revealing spatial correspondences with the pre-operative input. The framework employs an encoder for feature extraction and dual decoders for predicting spatial deformations and synthesizing the translated image. The predicted deformation field imposes spatial constraints on the generated output, ensuring anatomical consistency with the source. Extensive experiments on medical cosmetology and brain MRI datasets demonstrate that TraceTrans delivers accurate and interpretable post-operative predictions, highlighting its potential for reliable clinical deployment.

This paper combines vector-valued reproducing kernel Hilbert space (vRKHS) embedding with robust adaptive observation, yielding an algorithm that is both non-parametric and robust. The main contribution of this paper lies in the ability of the proposed system to estimate the state of a plan model whose matched uncertainties are elements of an infinite-dimensional native space. The plant model considered in this paper also suffers from unmatched uncertainties. Finally, the measured output is affected by disturbances as well. Upper bounds on the state observation error are provided in an analytical form. The proposed theoretical results are applied to the problem of estimating the state of a rigid body.

Creativity is a core cognitive capacity underlying innovation and adaptive problem solving, yet how it is represented in the brain's intrinsic functional architecture is not fully understood. While resting-state fMRI studies have identified large-scale network correlates associated with differences in creativity, EEG provides the temporal resolution for examining oscillatory dynamics contributing to intrinsic network organization. We examined whether resting-state EEG connectivity patterns are associated with individual differences across multiple creativity-related measures. Thirty healthy young adults completed a multidimensional creativity battery comprising the Inventory of Creative Activities and Achievements (ICAA), the Divergent Association Task (DAT), the Matchstick Arithmetic Puzzles Task (MAPT) and a Self-rating (SR) of creative ability. Graph-theoretical analyses of alpha-band functional connectivity revealed two participant groups, each with distinct patterns of neural activity: Cluster 1 showed reduced global connectivity with relatively preserved left frontal connectivity and greater network modularity; Cluster 0 exhibited stronger overall connectivity strength, reduced modularity and higher local clustering. Notably, Cluster 1 reported higher self-rated creative ability and more frequent engagement in real-world creative activities. These findings suggest that resting-state EEG connectivity patterns are associated with variation in creative self-efficacy and creative engagement, highlighting characteristic patterns of alpha-band network organization observed at rest.

The integration of converter-interfaced generation introduces new transient stability challenges to modern power systems. Classical Lyapunov- and scalable passivity-based approaches typically rely on restrictive assumptions, and finding storage functions for large grids is generally considered intractable. Furthermore, most methods require an accurate grid dynamics model. To address these challenges, we propose a model-free, nonlinear, and dissipativity-based controller which, when applied to grid-connected virtual synchronous generators (VSGs), enhances power system transient stability. Using input-state data, we train neural networks to learn dissipativity-characterizing matrices that yield stabilizing controllers. Furthermore, we incorporate cost function shaping to improve the performance with respect to the user-specified objectives. Numerical results on a modified, all-VSG Kundur two-area power system validate the effectiveness of the proposed approach.

Energy communities (ECs) are emerging as a promising decentralized model for managing cooperative distributed energy resources (DERs). As these communities expand and their operations become increasingly integrated into the grid, ensuring fairness in allocating operating costs among participants becomes a challenge. In distribution networks, DER operations at the community level can influence Distribution Locational Marginal Prices (DLMPs), which in turn affect system's operation cost. This interdependence between local decisions and system-level pricing introduces new challenges for fair and transparent cost allocation. Despite growing interest in fairness-aware methods, most methods do not account for the impact of DLMPs. To fill this gap, we propose a bilevel optimization model in which a Community Energy Aggregator (CEA) schedules DERs across multiple ECs while a Distribution System Operator (DSO) determines DLMPs through network-constrained dispatch. Leveraging the Karush-Kuhn-Tucker (KKT) conditions and strong duality, the bilevel model is reformulated into a tractable single-level problem. We achieve fairness in the cost allocation by applying the Shapley value to quantify each community's marginal contribution to system-wide cost savings. The effectiveness of the proposed method is validated through simulations on several benchmark distribution systems.

The introduction of Integrated Sensing and Communications (ISAC) in cellular systems is not expected to result in a shift away from the popular choice of cost- and energy-efficient analog or hybrid beamforming structures. However, this comes at the cost of limiting the angular capabilities to a confined space per acquisitions. Thus, as a prerequisite for the successful implementation of numerous ISAC use cases, the need for an optimal angular estimation of targets and their separation based on the minimal number of angular samples arises. In this work, different approaches for angular estimation based on a minimal, DFT-based set of angular samples are evaluated. The samples are acquired through sweeping multiple beams of an ISAC proof of concept (PoC) in the industrial scenario of the ARENA2036. The study's findings indicate that interpolation approaches are more effective for generalizing across different types of angular scenarios. While the orthogonal matching pursuit (OMP) approach exhibits the most accurate estimation for a single, strong and clearly discriminable target, the DFT-based interpolation approach demonstrates the best overall estimation performance.

Wide-bandgap (WBG) technologies offer unprecedented improvements in power system efficiency, size, and performance, but also introduce unique sensor corruption and cybersecurity risks in industrial control systems (ICS), particularly due to high-frequency noise and sophisticated cyber-physical threats. This proof-of-concept (PoC) study demonstrates the adaptation of a noise-driven physically unclonable function (PUF) and machine learning (ML)-assisted anomaly detection framework to the demanding environment of WBG-based ICS sensor pathways. By extracting entropy from unavoidable WBG switching noise (up to 100 kHz) as a PUF source, and simultaneously using this noise as a real-time threat indicator, the proposed system unites hardware-level authentication and anomaly detection. Our approach integrates hybrid machine learning (ML) models with adaptive Bayesian filtering, providing robust and low-latency detection capabilities resilient to both natural electromagnetic interference (EMI) and active adversarial manipulation. Through detailed simulations of WBG modules under benign and attack scenarios–including EMI injection, signal tampering, and node impersonation–we achieve 95% detection accuracy and sub-millisecond processing latency. These results demonstrate the feasibility of physics-driven, dual-use noise exploitation as a scalable ICS defense primitive. Our findings lay the groundwork for next-generation security strategies that leverage inherent device characteristics, bridging hardware and artificial intelligence (AI) for enhanced protection of critical ICS infrastructure.

This paper aims to investigate the distributed stochastic optimization problems on compact embedded submanifolds (in the Euclidean space) where the local cost functions are weakly-convex. To address the manifold structure, we propose a distributed Riemannian stochastic proximal algorithm framework by utilizing the retraction and Riemannian consensus protocol, and analyze three specific algorithms: the distributed Riemannian stochastic subgradient, proximal point, and prox-linear algorithms. When the initial points satisfy certain conditions, we show that the iterates generated by this framework converge to a nearly stationary point in expectation while achieving consensus. We further establish the convergence rate of the algorithm framework as $\mathcal{O}(\frac{1+\kappa_g}{\sqrt{k}})$ where $k$ denotes the number of iterations and $\kappa_g$ shows the impact of manifold geometry on the algorithm performance. Finally, numerical experiments are implemented to demonstrate the theoretical results and show the empirical performance.

Multimodal respiratory sound classification offers promise for early pulmonary disease detection by integrating bioacoustic signals with patient metadata. Nevertheless, current approaches remain vulnerable to spurious correlations from attributes such as age, sex, or acquisition device, which hinder their generalization, especially under distribution shifts across clinical sites. To this end, we propose a counterfactual adversarial debiasing framework. First, we employ a causal graph-based counterfactual debiasing methodology to suppress non-causal dependencies from patient metadata. Second, we introduce adversarial debiasing to learn metadata-insensitive representations and reduce metadata-specific biases. Third, we design counterfactual metadata augmentation to mitigate spurious correlations further and strengthen metadata-invariant representations. By doing so, our method consistently outperforms strong baselines in evaluations under both in-distribution and distribution shifts. Code is available at https://github.com/RSC-Toolkit/BTS-CARD.

This paper investigates the performance of Binaural Signal Matching (BSM) methods for near-field sound reproduction using a wearable glasses-mounted microphone array. BSM is a flexible, signal-independent approach for binaural rendering with arbitrary arrays, but its conventional formulation assumes far-field sources. In our previous work, we proposed a near-field extension of BSM (NF-BSM) that incorporates distance-dependent modeling and showed improved performance over far-field BSM using analytic data, though degradation persisted for sources very close to the array. In this study, we extend that analysis by using realistic simulated data of near-field Head-Related Transfer Functions (HRTFs) and Acoustic Transfer Functions (ATFs) of the array, accounting for listener head rotation and evaluating binaural cues such as interaural level and time differences (ILD and ITD). A key contribution is the introduction of a Field of View (FoV) weighting, designed to emphasize perceptually relevant directions and improve robustness under challenging conditions. Results from both simulation and a listening test confirm that NF-BSM outperforms traditional far-field BSM in near-field scenarios, and that the proposed NF-FoV-BSM method achieves the best perceptual and objective quality among all tested methods, particularly at close source distances and under head rotation. These findings highlight the limitations for far-field models in near-field sources and demonstrate that incorporating source distance and directional weighting can significantly improve binaural reproduction performance for wearable spatial audio systems.

Chromatin sensitive partial wave spectroscopic (csPWS) microscopy enables label free detection of nanoscale chromatin packing alterations that occur before visible cellular transformation. However, manual nuclear segmentation limits population scale analysis needed for biomarker discovery in early cancer detection. The lack of annotated csPWS imaging data prevents direct use of standard deep learning methods. We present CFU Net, a hierarchical segmentation architecture trained with a three stage curriculum on synthetic multimodal data. CFU Net achieves near perfect performance on held out synthetic test data that represent diverse spectroscopic imaging conditions without manual annotations (Dice 0.9879, IoU 0.9895). Our approach uses physics based rendering that incorporates empirically supported chromatin packing statistics, Mie scattering models, and modality specific noise, combined with a curriculum that progresses from adversarial RGB pretraining to spectroscopic fine tuning and histology validation. CFU Net integrates five architectural elements (ConvNeXt backbone, Feature Pyramid Network, UNet plus plus dense connections, dual attention, and deep supervision) that together improve Dice over a baseline UNet by 8.3 percent. We demonstrate deployment ready INT8 quantization with 74.9 percent compression and 0.15 second inference, giving a 240 times throughput gain over manual analysis. Applied to more than ten thousand automatically segmented nuclei from synthetic test data, the pipeline extracts chromatin biomarkers that distinguish normal from pre cancerous tissue with large effect sizes (Cohens d between 1.31 and 2.98), reaching 94 percent classification accuracy. This work provides a general framework for synthetic to real transfer learning in specialized microscopy and open resources for community validation on clinical specimens.

Automated dysarthria detection and severity assessment from speech have attracted significant research attention due to their potential clinical impact. Despite rapid progress in acoustic modeling and deep learning, models still fall short of human expert performance. This manuscript provides a comprehensive analysis of the reasons behind this gap, emphasizing a conceptual divergence we term the “perceptual-statistical gap”. We detail human expert perceptual processes, survey machine learning representations and methods, review existing literature on feature sets and modeling strategies, and present a theoretical analysis of limits imposed by label noise and inter-rater variability. We further outline practical strategies to narrow the gap, perceptually motivated features, self-supervised pretraining, ASR-informed objectives, multimodal fusion, human-in-the-loop training, and explainability methods. Finally, we propose experimental protocols and evaluation metrics aligned with clinical goals to guide future research toward clinically reliable and interpretable dysarthria assessment tools.

Channel estimation for massive multiple-input multiple-output (MIMO) systems is fundamentally constrained by excessive pilot overhead and high estimation latency. To overcome these obstacles, recent studies have leveraged deep generative networks to capture the prior distribution of wireless channels. In this paper, we propose a novel estimation framework that integrates an energy-based generative diffusion model (DM) with the Metropolis-Hastings (MH) principle. By reparameterizing the diffusion process with an incorporated energy function, the framework explicitly estimates the unnormalized log-prior, while MH corrections refine the sampling trajectory, mitigate deviations, and enhance robustness, ultimately enabling accurate posterior sampling for high-fidelity channel estimation. Numerical results reveal that the proposed approach significantly improves estimation accuracy compared with conventional parameterized DMs and other baseline methods, particularly in cases with limited pilot overhead.

The integration of Artificial Intelligence (AI) into safety-critical systems introduces a new reliability paradigm: silent failures, where AI produces confident but incorrect outputs that can be dangerous. This paper introduces the Formal Assurance and Monitoring Environment (FAME), a novel framework that confronts this challenge. FAME synergizes the mathematical rigor of offline formal synthesis with the vigilance of online runtime monitoring to create a verifiable safety net around opaque AI components. We demonstrate its efficacy in an autonomous vehicle perception system, where FAME successfully detected 93.5% of critical safety violations that were otherwise silent. By contextualizing our framework within the ISO 26262 and ISO/PAS 8800 standards, we provide reliability engineers with a practical, certifiable pathway for deploying trustworthy AI. FAME represents a crucial shift from accepting probabilistic performance to enforcing provable safety in next-generation systems.

This work presents a unified framework for estimating both sound-field direction and diffuseness using practical microphone arrays with different spatial configurations. Building on covariance-based diffuseness models, we formulate a velocity-only covariance approach that enables consistent diffuseness evaluation across heterogeneous array geometries without requiring mode whitening or spherical-harmonic decomposition. Three array types – an A-format array, a rigid-sphere array, and a newly proposed tight-frame array – are modeled and compared through both simulations and measurement-based experiments. The results show that the tight-frame configuration achieves near-isotropic directional sampling and reproduces diffuseness characteristics comparable to those of higher-order spherical arrays, while maintaining a compact physical structure. We further examine the accuracy of direction-of-arrival estimation based on acoustic intensity within the same framework. These findings connect theoretical diffuseness analysis with implementable array designs and support the development of robust, broadband methods for spatial-sound-field characterization.

For a wide range of envisioned integrated sensing and communication (ISAC) use cases, it is necessary to incorporate tracking techniques into cellular communication systems. While numerous multi-target tracking (MTT) algorithms exist, they have not yet been applied to real-world ISAC, with its challenges such as clutter and non-optimal hardware with design emphasis on communication instead of sensing. In this work, we showcase MTT based on the probability hypothesis density (PHD) filter in the range and radial speed domain. The measurements are taken with a 5G compliant ISAC proof-of-concept in a real factory environment, where the pedestrian-like targets are generated by a radar target emulator. We detail the complete pipeline, from measurement acquisition to evaluation, with a focus on the post-processing of the raw captured data and the tracking itself. Our end-to-end evaluation and comparison to simulations show good MTT performance with mean absolute ranging error <1.5m and detection rates >91% for realistic but challenging scenarios.

Machine learning in neurosurgery is limited by challenges in assembling large, high-quality imaging datasets. Synthetic data offers a scalable, privacy-preserving solution. We evaluated the feasibility of generating realistic lateral cervical spine radiographs using a denoising diffusion probabilistic model (DDPM) trained on 4,963 images from the Cervical Spine X-ray Atlas. Model performance was monitored via training/validation loss and Frechet inception distance, and synthetic image quality was assessed in a blinded "clinical Turing test" with six neuroradiologists and two spine-fellowship trained neurosurgeons. Experts reviewed 50 quartets containing one real and three synthetic images, identifying the real image and rating realism on a 4-point Likert scale. Experts correctly identified the real image in 29% of trials (Fleiss' kappa=0.061). Mean realism scores were comparable between real (3.323) and synthetic images (3.228, 3.258, and 3.320; p=0.383, 0.471, 1.000). Nearest-neighbor analysis found no evidence of memorization. We also provide a dataset of 20,063 synthetic radiographs. These results demonstrate that DDPM-generated cervical spine X-rays are statistically indistinguishable in realism and quality from real clinical images, offering a novel approach to creating large-scale neuroimaging datasets for ML applications in landmarking, segmentation, and classification.

Low-light image enhancement (LLIE) aims at improving the perception or interpretability of an image captured in an environment with poor illumination. With the advent of deep learning, the LLIE technique has achieved significant breakthroughs. However, existing LLIE methods either ignore the important role of frequency domain information or fail to effectively promote the propagation and flow of information, limiting the LLIE performance. In this paper, we develop a novel frequency-spatial interaction-driven network (FSIDNet) for LLIE based on two-stage architecture. To be specific, the first stage is designed to restore the amplitude of low-light images to improve the lightness, and the second stage devotes to restore phase information to refine fine-grained structures. Considering that Frequency domain and spatial domain information are complementary and both favorable for LLIE, we further develop two frequency-spatial interaction blocks which mutually amalgamate the complementary spatial and frequency information to enhance the capability of the model. In addition, we construct the Information Exchange Module (IEM) to associate two stages by adequately incorporating cross-stage and cross-scale features to effectively promote the propagation and flow of information in the two-stage network structure. Finally, we conduct experiments on several widely used benchmark datasets (i.e., LOL-Real, LSRW-Huawei, etc.), which demonstrate that our method achieves the excellent performance in terms of visual results and quantitative metrics while preserving good model efficiency.

Vision-Language Models (VLMs) have shown significant progress in open-set challenges. However, the limited availability of 3D datasets hinders their effective application in 3D scene understanding. We propose LOC, a general language-guided framework adaptable to various occupancy networks, supporting both supervised and self-supervised learning paradigms. For self-supervised tasks, we employ a strategy that fuses multi-frame LiDAR points for dynamic/static scenes, using Poisson reconstruction to fill voids, and assigning semantics to voxels via K-Nearest Neighbor (KNN) to obtain comprehensive voxel representations. To mitigate feature over-homogenization caused by direct high-dimensional feature distillation, we introduce Densely Contrastive Learning (DCL). DCL leverages dense voxel semantic information and predefined textual prompts. This efficiently enhances open-set recognition without dense pixel-level supervision, and our framework can also leverage existing ground truth to further improve performance. Our model predicts dense voxel features embedded in the CLIP feature space, integrating textual and image pixel information, and classifies based on text and semantic similarity. Experiments on the nuScenes dataset demonstrate the method's superior performance, achieving high-precision predictions for known classes and distinguishing unknown classes without additional training data.

Modeling dynamical systems is crucial across the science and engineering fields for accurate prediction, control, and decision-making. Recently, machine learning (ML) approaches, particularly neural ordinary differential equations (NODEs), have emerged as a powerful tool for data-driven modeling of continuous-time dynamics. Nevertheless, standard NODEs require a large number of data samples to remain consistent under varying control inputs, posing challenges to generate sufficient simulated data and ensure the safety of control design. To address this gap, we propose trajectory-sensitivity-aware (TRASE-)NODEs, which construct an augmented system for both state and sensitivity, enabling simultaneous learning of their dynamics. This formulation allows the adjoint method to update gradients in a memory-efficient manner and ensures that time-invariant control set-point effects are captured in the learned dynamics. We evaluate TRASE-NODEs using damped oscillator and inverter-based resources (IBRs). The results show that TRASE-NODEs generalize better from the limited training data, yielding lower prediction errors than standard NODEs for both examples. The proposed framework offers a data-efficient, control-oriented modeling approach suitable for dynamic systems that require accurate trajectory sensitivity prediction.

The paper extends core results of behavioral systems theory from linear to affine time-invariant systems. We characterize the behavior of affine time-invariant systems via kernel, input-output, state-space, and finite-horizon data-driven representations, demonstrating a range of structural parallels with linear time-invariant systems. Building on these representations, we introduce a new persistence of excitation condition tailored to the model class of affine time-invariant systems. The condition yields a new fundamental lemma that parallels the classical result for linear systems while provably reducing data requirements. Our analysis highlights that excitation conditions must be adapted to the model class: overlooking structural differences may lead to unnecessarily conservative data requirements.

This chapter bridges technical analysis and organizational preparedness by tracing the path from layered failure modes to reliability awareness in generative and agentic AI systems. We first introduce an 11-layer failure stack, a structured framework for identifying vulnerabilities ranging from hardware and power foundations to adaptive learning and agentic reasoning. Building on this, the chapter demonstrates how failures rarely occur in isolation but propagate across layers, creating cascading effects with systemic consequences. To complement this diagnostic lens, we develop the concept of awareness mapping: a maturity-oriented framework that quantifies how well individuals and organizations recognize reliability risks across the AI stack. Awareness is treated not only as a diagnostic score but also as a strategic input for AI governance, guiding improvement and resilience planning. By linking layered failures to awareness levels and further integrating this into Dependability-Centred Asset Management (DCAM), the chapter positions awareness mapping as both a measurement tool and a roadmap for trustworthy and sustainable AI deployment across mission-critical domains.

The Koopman operator and extended dynamic mode decomposition (EDMD) as a data-driven technique for its approximation have attracted considerable attention as a key tool for modeling, analysis, and control of complex dynamical systems. However, extensions towards control-affine systems resulting in bilinear surrogate models are prone to demanding data requirements rendering their applicability intricate. In this paper, we propose a framework for data-fitting of control-affine mappings to increase the robustness margin in the associated system identification problem and, thus, to provide reliable bilinear EDMD schemes. In particular, guidelines for input selection based on subspace angles are deduced such that a desired threshold with respect to the minimal singular value is ensured. Moreover, we derive necessary and sufficient conditions of optimality for maximizing the minimal singular value. Further, we demonstrate the usefulness of the proposed approach using bilinear EDMD with control for nonholonomic robots.

In this study, we present a deep learning framework that learns complex spatio-temporal correlation structures of EEG signals through a Spatio-Temporal Attention Network (STAN) for accurate predictions of onset of seizures for Epilepsy patients. Unlike existing methods, which rely on feature engineering and/or assume fixed preictal durations, our approach simultaneously models spatio-temporal correlations through STAN and employs an adversarial discriminator to distinguish preictal from interictal attention patterns, enabling patient-specific learning. Evaluation on CHB-MIT and MSSM datasets demonstrates 96.6% sensitivity with 0.011/h false detection rate on CHB-MIT, and 94.2% sensitivity with 0.063/h FDR on MSSM, significantly outperforming state-of-the-art methods. The framework reliably detects preictal states at least 15 minutes before an onset, with patient-specific windows extending to 45 minutes, providing sufficient intervention time for clinical applications.

This work studies two types of computer networking models. The primary focus is to understand the different dynamical phenomena observed in practice due to the presence of severe nonlinearities, delays and widely varying operating conditions. The first models considered are of senders running TCP (Transmission Control Protocol) and traffic passing through RED (Random Early Detection) gateways. Building on earlier work, a first order nonlinear discrete-time model is developed for the interaction scenario between transport protocols like TCP and UDP (User Datagram Protocol) and Active Queuing Management schemes like RED. It is shown that the dynamics resulting from the interaction with TCP is consistent with various dynamical behaviors and parameter sensitivities observed in practice. Using bifurcation-theoretic ideas it is shown that TCP-RED type networks may lose their stability through a period doubling bifurcation followed by border collision bifurcations. The nonlinear dependence of the throughput function of TCP-type flows on drop probability is found to be responsible for the period doubling bifurcation, whereas limited buffer space and lack of sufficient damping results in border collision bifurcations. A second class of models studied in this work deals with optimal rate control in networks and are based on the rate-control framework proposed by Kelly. Using the results on delay-differential equation stability, the stability and its lack thereof is studied through an underlying map which arises naturally in time delay systems. An invariance property of this map is used to prove delay-independent stability and to compute bounds on periodic oscillations.