This work introduces a nonlinear system identification method that uses frequency response data to estimate a compact and interpretable nonlinear state-space model of structural systems. A key novelty is the ability to directly exploit data from experimental continuation, including unstable branches, by minimizing the discrepancy between measured and predicted outputs across multiple harmonics. This enables accurate characterization of nonlinear dynamics from a limited dataset, while keeping computational costs low and ensuring robustness to noise and initialization. The method, called NFR-ID, is validated through two experimental cases: an electronic Duffing oscillator and a thin-walled plate, the latter exhibiting rich nonlinear behavior due to large-amplitude vibrations. Results demonstrate the accuracy and broad applicability of NFR-ID as an efficient and general framework for analyzing nonlinear structural dynamics.

Abstract Comparative genomics enables scientific exploration of genetic variations across species, enabling insights into human diseases that are indiscernible from analysis of human data alone. Dogs, in particular, are an excellent model organism in comparative oncology because they develop spontaneous tumors that closely resemble human cancers. Many recent studies have investigated parallels between dog and human cancers in a single tumor type. However, to date, no study has systematically quantified the effectiveness of the dog model as a pan-cancer representation of adult and pediatric cancers. To address this gap, we performed a pan-cancer RNA-seq analysis of human and dog cancers spanning 6,048 samples and 11 tumor types, including 987 dog tumor samples from Cahill et al., 4,272 adult samples from TCGA, and 789 pediatric samples from the Treehouse compendium. We analyzed 10 adult and 5 pediatric cancers with at least 10 tumor samples for each species. Using a combination of exploratory data analysis, differential expression and enrichment analysis, unsupervised clustering, and supervised modeling, we identified shared expression patterns across species. We developed a weighted formula incorporating results from these analyses to quantitatively assess the translational relevance of the dog model for each cancer and applied it across cancer types. Our formula ranked dog cancers by transcriptional similarity to human cancers and revealed which models best capture tumor heterogeneity. The cancer-type specific results from this study match those from individual cancer studies, suggesting that our formula effectively identifies the tumor types where dog models best reflect human cancers. Our work introduces a unified, quantitative framework for evaluating cross-species cancer models through combining multiple analysis methods to support preclinical studies and accelerate therapy development. This scalable approach enables fast and accurate identification of model systems to study rare human cancers. As genome sequencing increases, this automated approach minimizes the manual effort required to identify suitable model organisms across the tree of life and opens new opportunities for comparative oncology studies. Citation Format: Geesa Daluwatumulle, Leslie Smith, Nathan Glen, James Cahill, Kiley Graim. Quantifying the translational relevance of naturally occurring dog cancers as models of adult and pediatric tumors [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Cancer Evolution: The Dynamics of Progression and Persistence; 2025 Dec 4-6; Albuquerque, NM. Philadelphia (PA): AACR; Cancer Res 2025;85(23_Suppl):Abstract nr B029.

A sparse extended Kalman filtering based on Laplace prior for real-time state-parameter identification of time-varying structural systems

Unbalance identification of high-speed aerostatic conical bearing-rotor system based on Bayesian adaptive optimization CNN-Res time-domain automatic feature extraction algorithm

Asymptotically consistent identification of FIR systems with binary-valued observations subject to sequence-type replay attacks

Normalized SPSA for Hammerstein model identification of twin rotor and electro-mechanical positioning systems

Model-free system identification of surface ships in waves via Hankel dynamic mode decomposition with control

Data-driven modeling and prediction of microglial cell dynamics in the ischemic penumbra.

A state alignment-centric approach to federated system identification: The FedAlign framework

This research investigates the identification problem of fuzzy systems represented by fuzzy relational equations and TSK-type fuzzy models under uncertainty. The research object is the nonlinear dynamic model of a steam generator of a thermal power plant, for which accurate modeling is essential due to its complex behavior. The scientific problem addressed in the article is determining the optimal fuzzy implication and developing an identification algorithm that minimizes modeling error for nonlinear technological objects. An identification approach based on max–min composition is constructed using a fuzzy rule base to model input–output relationships. Structural and parametric identification procedures are formulated to select the criteria, parameters, and structural components of the fuzzy difference model. Several nonlinear control algorithms and multiple implication types are tested on the steam generator model. Experimental analysis shows that the ALI1 implication achieves the minimum mean square error among the evaluated implications, providing more accurate fuzzy relational mapping. The obtained results improve the quality of fuzzy system identification and enable the synthesis of an efficient fuzzy control strategy for nonlinear industrial processes. The developed method can be practically applied in real-time modeling, control, and optimization of thermal power plant units

Abstract The martensitic transformation in shape memory alloys involves the formation of twinned martensite regions at the austenite-martensite phase boundaries. The twin boundary energy plays a significant role in determining the twinned microstructure, as well as the hysteresis and kinetics of the transformation. We present a systematic microscopy-based approach to analyze the microstructure of twinned martensite plates, enabling the identification of the twin system and extraction of the energies of twin boundaries and martensite plates. The method is applied to the abundant ⟨011⟩ type II twin boundary in the B-II-1 system, in NiTi. The obtained knowledge paves the route for modeling the macroscale evolution of the phase transformation based on material properties measured at the nanoscale.

The complex time addiction and randomness of multivariate time series make it nec-essary to apply time series analysis to these variables. So, it is important to produce methods that can be used appropriately for time series system identification. Time series are generally handled as a single-layer architecture consisting of only the observed data processing layer. In this study, a hybrid method has three-layer adapted and evaluated using an adaptive neuro-fuzzy inference system (ANFIS). The main motivation of this study is to learn the errors produced by ANFIS method and to use them as new in-formation. For this purpose, proposed method was evaluated using real-time serial data sets obtained from different fields. Due to the three-layer architecture, the errors caused by the results produced by ANFIS are reused. As a result, the method of learning from errors has been realized and better results have been produced com-pared to traditional single-layer architecture.

Cross-electrophile coupling (XEC) reactions have emerged as powerful methods for carbon-carbon bond formation and have seen rapid development in recent years. Among these, nickel-catalyzed XEC offers a cost-effective and versatile alternative to palladium-based biaryl synthesis, gaining increasing prominence in organic synthesis. However, despite the advantages of nickel-based systems, the identification of optimal catalytic systems and a comprehensive understanding of the reaction mechanisms involved remain limited due to the complex nature of Ni complexes. Electrochemistry provides a compelling platform for studying nickel-catalyzed reductive coupling reactions. It allows for precise control of redox potentials, eliminates the need for metallic reductants, decouples electrochemical and chemical processes, and enhances scalability, all of which have spurred growing interest in electrochemical approaches. Moreover, electrochemical techniques offer unique mechanistic insights into nickel-mediated transformations.[1] The generally accepted mechanism for reductive nickel-catalyzed coupling of aryl halides involves: (a) reduction of a Ni²⁺ precursor to Ni⁰, (b) oxidative addition of the aryl halide, (c) transmetalation, and (d) reductive elimination. Many of the reactants and intermediates in this catalytic cycle exhibit well-defined redox behavior. Owing to the distinct redox characteristics of these species, electroanalytical methods are well-suited for monitoring concentration profiles and enabling kinetic investigations during the reaction.[2] In this study, we employ cyclic voltammetry (CV) to explore the redox behavior of nickel complexes relevant to XEC reactions. CV provides information on thermodynamic potentials and allows quantification of both the relative and absolute concentrations of nickel species throughout the catalytic cycle. We investigate how variables such as temperature, solvent, ligand electronics, and substrate choice affect nickel speciation and reactivity. In particular, voltammetric techniques are applied to determine the redox potentials, stability, and reactivity of intermediates, especially those formed during the oxidative addition of Ni to aryl and alkyl halides, based on the analysis of voltammetric peak currents and potentials. These data offer valuable kinetic and mechanistic insights into nickel-catalyzed cross-electrophile coupling reactions.[1] M. Rafiee, D.J. Abrams, L. Cardinale, Z. Goss, A. Romero-Arenas, S.S. Stahl, Cyclic Voltammetry And Chronoamperometry: Mechanistic Tools for Organic Electrosynthesis. Chem. Soc. Rev. 2024, 53, 566.[2] Z.M. Su, J. Zhu, D.L. Poole, M. Rafiee, R.S. Paton, D.J. Weix, S.S. Stahl, Selective Ni-Catalyzed Cross-Electrophile Coupling of Heteroaryl Chlorides and Aryl Bromides at 1: 1 Substrate Ratio. J. Am. Chem. Soc. 2025, 147, 353.

Abstract A central challenge in the control of complex industrial processes like heat exchangers is the persistent disconnection between global offline optimization and real-time online adaptation, leading to suboptimal performance under dynamic conditions. To address this, this paper proposes a coordinated hierarchical control strategy, CB-BPP, a three-layer framework integrating Chaotic Particle Swarm Optimization (CFPSO), a Backpropagation Neural Network (BPNN) identifier, and a Predictive Functional Control (PFC)-based PID controller. At the top layer, CFPSO performs offline global optimization of the BPNN's initial weights and establishes safe bounds for controller parameters. The middle layer employs the BPNN for online system identification and real-time adaptation of PID parameters. The bottom layer executes high-precision control. This tiered architecture decouples offline optimization from online adaptation, enabling synergistic control. Comprehensive simulations demonstrate the superiority of CB-BPP, reducing the Integral of Absolute Error (IAE) by up to 45% and improving recovery speed by 35% compared with advanced baseline methods. Furthermore, extensive robustness analysis against sensor noise and parameter drift, alongside a successful Hardware-in-the-Loop (HIL) implementation, validates its practical applicability. The results confirm the proposed method's high precision, strong robustness, and real-time feasibility, providing an advanced control solution for complex industrial processes.

Artificial Intelligence (AI) is rapidly transforming engineering fields, from robotics to aerospace, with applications in control systems for UAVs and satellites. This work builds on a previously developed AI attitude controller for the InnoCube 3U nanosatellite. Deploying complex Neural Networks (NNs) on resource-limited microcontrollers presents a significant challenge. To overcome this, we propose distilling a Multi-Layer Perceptron (MLP) trained with Deep Reinforcement Learning (DRL) for attitude control into a Kolmogorov–Arnold Network (KAN). We convert this numeric KAN into a symbolic KAN, where each edge represents a learnable mathematical function, and finally extract a concise symbolic formula. This symbolic representation dramatically reduces memory usage and computational complexity, making it ideal for pico- and nanosatellites. We evaluate and demonstrate the feasibility of this approach for inertial pointing with reaction wheels in simulation using a realistic model of the InnoCube satellite. Our results show that the highly compressed KANs successfully solve the attitude control problem, while reducing the required memory footprint and inference time on the InnoCube ADCS hardware by over an order of magnitude. Beyond attitude control, we believe symbolic KANs hold great potential in aerospace for neural network compression and interpretable, data-driven modeling and system identification in future space missions.

Modeling the complex nonlinear dynamics of Brushless DC motors has been a prominent research focus over the past two decades, driven by their superior advantages and widespread industrial applications. Despite extensive efforts, achieving high-efficiency prediction of speed and torque responses remains a challenge. This study proposes a hybrid machine learning-based approach using the Nonlinear Autoregressive Neural Network with Exogenous Inputs. The method combines artificial neural networks and system identification techniques to enhance predictive accuracy in nonlinear dynamic systems. For both speed and torque modeling, optimal time delays and neural network layer sizes are selected to accurately capture the ripple effects under a multi-step input signal applied to a three-phase inverter. The proposed models yield Mean Square Error values as low as for speed and for torque. Regression coefficients of 1.000 for speed and 0.998 for torque are achieved consistently across training, validation, testing, and additional testing phases, following a data split of 70% for training and 15% each for validation and testing. To further evaluate generalization, the approach is tested using a distinct multi-step input voltage signal, with the results confirming the robustness and superiority of the proposed method in both speed and torque prediction. Comparative analysis with existing literature demonstrates the dominance of the proposed models. These high-fidelity models can serve as a foundation for designing advanced controllers aimed at efficient speed regulation and torque ripple mitigation in Brushless DC motors.

Many studies on acoustic radiation forces, especially those applied to acoustic levitation, focus on characterizing the behaviour of acoustic fields. However, the dynamic response of the levitated objects, particularly those larger than the wavelength limit, remains relatively underexplored. Here, we look to bridge this gap by deriving nonlinear equations of motion for a spherical object trapped under acoustic radiation forces while subject to external excitation. For such a contemporary scenario, the otherwise elemental Gorkov formulation fails to provide accurate results. Using Sparse Identification of Nonlinear Dynamical Systems (SINDy), first, we derive the corresponding nonlinear equation of motion from analytical time series data obtained through the Gorkov formulation and external excitation for acoustically small objects. This approach recovers the governing equation with less than 0.05% error in coefficient values when compared to the analytical solution. Second, we conduct experiments with the TinyLev levitator with external excitation applied via an external actuator to generate the required time series for an acoustically large object. SINDy is applied to reconstruct governing equations from experimental data, allowing for the study of how excitation amplitude affects acoustically large objects. All obtained coefficients change with excitation amplitude, and the coefficients in the dynamic equation of motion should not be treated as constants. Strong velocity-dependent terms emerged, indicating a complex relationship between viscosity and object response, which classical models do not predict. The bifurcation diagram obtained using the SINDy-derived equation of motion shows closer agreement with that obtained experimentally. These results demonstrate that SINDy can recover equations consistent with Gorkov's formulation and extend beyond it, providing a pathway to derive analytical expressions directly from data for levitating and manipulating objects beyond the Rayleigh limit.

Biological systems inherently exhibit multi-scale dynamics, making accurate system identification particularly challenging due to the complexity of capturing a wide time scale spectrum. Traditional methods capable of addressing this issue rely on explicit equations, limiting their applicability in cases where only observational data are available. To overcome this limitation, we propose a data-driven framework that integrates the Sparse Identification of Nonlinear Dynamics (SINDy) method, the multi scale analysis algorithm Computational Singular Perturbation (CSP) and neural networks (NNs). This framework allows the partition of the available dataset in subsets characterized by similar dynamics, so that system identification can proceed within these subsets without facing a wide time scale spectrum. Accordingly, when the full dataset does not allow SINDy to identify the proper model, CSP is employed for the generation of subsets of similar dynamics, which are then fed into SINDy. CSP requires the availability of the gradient of the vector field, which is estimated by the NNs. The framework is tested on the Michaelis-Menten model, for which various reduced models in analytic form exist at different parts of the phase space. It is demonstrated that the CSP-based data subsets allow SINDy to identify the proper reduced model in cases where the full dataset does not. In addition, it is demonstrated that the framework succeeds even in the cases where the available data set originates from stochastic versions of the Michaelis-Menten model. This framework is algorithmic, so system identification is not hindered by the dimensions of the dataset.

Neurons in the early visual cortex respond selectively to multiple features of visual stimuli, but they respond inconsistently to repeated presentation of the same visual stimulus. Such trial-to-trial response variabilities are often treated as random noise and addressed by simple trial-averaging to obtain the stimulus-driven response. However, response variability may primarily be caused by non-sensory factors, particularly by variations in cortical state. Here we recorded and analyzed neuronal spiking activity in response to natural images from areas 17 and 18 of cats, along with local population neuronal signals, i.e., local field potentials (LFPs) and multi-unit activity (MUA). Single neurons showed highly varying degrees of trial-to-trial response variability, even when recorded simultaneously. We used a variability ratio (VR) measure to quantify the trial-wise differences in neural responses, and two cortical state indicative measures, a global fluctuation index (GFI) calculated using MUA, and a synchrony index (SI) calculated from LFP signals. We propose a compact convolutional neural network model with parallel pathways, to capture the stimulus-driven activity and the cortical state-driven response variabilities. The stimulus-driven pathway is comprised of a spatiotemporal filter, a parametric rectifier and a Gaussian map, and the cortical state-driven pathway contains temporal filters for MUA and LFPs. The model parameters are fit to best predict each neuron's spiking activity. We further evaluated the improvements in estimated receptive fields of neurons when incorporating cortical state related information in our system identification model. The fitted model performed with a significantly higher accuracy in predicting neural responses as well as qualitative improvements in the estimated receptive fields compared to a basic model with a stimulus-driven pathway alone. The neurons with higher response variability benefited more from the cortical state-driven pathway compared to less variable neurons. These results show that different neurons may differ greatly in their variability and in the degree of their relationship to indicators of cortical state fluctuations.

ABSTRACT This paper presents an adaptive control framework for dual‐system VTOL UAVs capable of operating in both rotary‐wing and fixed‐wing modes. These aerial vehicles present considerable control challenges due to their nonlinear, time‐varying dynamics and inherent instability during flight‐mode transitions. The proposed approach addresses these issues by leveraging nonlinear system identification via Adaptive Sparse Identification of Nonlinear Dynamics (ASINDy) with a Lyapunov‐based Model Predictive Control (LMPC) scheme. This integrated framework facilitates continuous model updating and guarantees stable trajectory tracking and robust performance. Compared to the GA‐PID, the ASINDy–LMPC approach reduced tracking error by approximately 65%, maximum deviation by 67%, average deviation by 79%, and power consumption by 73% in simulation, while nearly halving the control effort. Preliminary hardware trials on a VTOL UAV prototype corroborate these trends, demonstrating consistent improvements during hovering and outdoor flights.