Search Results (764)

In this paper, we study the nonconvex quadratic programming over the intersection of two balls in n-dimensional complex space, which is called the uniform CDT subproblem and is significant in both optimization theory and applications. We first prove the hidden convexity of the problem by using the S-lemma. In order to construct an algorithm, we prove the hidden convexity again by reformulating it as a convex problem (C). Subsequently, we employ the eigenvalue approximation method and the generalized Nesterov’s accelerated gradient descent algorithm to solve the problem (C) within an error tolerance of $ϵ$. Then, we prove that the proposed algorithm is linear-time with respect to the density of the matrix in the objective function. The density is related to the number of non-zero entries of the matrix. We do numerical experiments to compare the proposed linear-time algorithm with the CVX solver, with the method proposed by Burer and with the algorithm proposed by Sakaue et al. Numerical results show that the linear-time algorithm performs much better than other methods, which verify that the proposed algorithm is applicable to many large-scale problems nowadays. Full article

To address the challenges of the large-scale georeferencing of tower-based cameras and the limited capability of video-based spatial analysis, we proposed a geospatial mapping method integrating 3D GIS and gradient descent optimization. Using a Digital Elevation Model (DEM), high-resolution remote sensing imagery, and tower-based video data as the primary data sources, the proposed method first estimates the intrinsic parameters of the tower-based camera by aligning a 3D GIS virtual camera with the video imagery. Subsequently, the initial camera extrinsic parameters are estimated using the PnP algorithm based on the previously estimated intrinsic matrix $K$ and the corresponding control point pairs. Building upon these initial estimates, the camera intrinsic and extrinsic parameters are jointly optimized using a constrained L-BFGS-B framework that incorporates prior knowledge of the tower planar location, explicit box constraints, and a semi-constrained parameterization scheme with bounded parameter ranges. Furthermore, an outlier-removal and re-optimization strategy is employed to further improve the accuracy of parameter estimation. Finally, the optimized parameters are employed to transform image coordinates into three-dimensional world coordinates, and video geospatial mapping is achieved through the integration of colored point clouds with the 3D GIS scene. The results showed the following: (1) The 3D GIS scene constructed from publicly available DEM and high-resolution remote sensing imagery met the requirements for the initial estimation of intrinsic and extrinsic camera parameters. (2) Compared with PnP, RANSAC-PnP, SQPnP, and DLT, the proposed method achieves lower reprojection and 3D spatial errors. For the independent check points, the RMSE of the reprojection error is reduced by 66.4%, 73.6%, 68.0%, and 48.3%, respectively, while the RMSE of the 3D spatial error is reduced by 84.6%, 86.2%, 83.1%, and 69.4%, respectively. These results demonstrate that the proposed method provides reliable camera parameter estimates for video geospatial mapping. (3) Using the estimated camera parameters, image coordinates are transformed into 3D world coordinates to generate a georeferenced colored point cloud, which facilitates integrated analysis with existing geospatial datasets. The proposed method provides a feasible solution for tower-based camera georeferencing and three-dimensional visualization under conditions without field calibration. It offers a theoretical and technical basis for geospatial monitoring and related applications. Full article

This paper proposes a low-complexity residual-corrected loss-minimization current-reference generation method for permanent magnet synchronous motor (PMSM) drives. Existing loss-minimization control (LMC) methods often rely on lookup tables, numerical optimization, high-order algebraic equations, or explicit approximations based on maximum torque per ampere (MTPA) and maximum torque per voltage (MTPV) references. While effective, these simplified approximations often introduce residual errors relative to the true loss-optimal conditions. To address this, this paper adopts a more consistent iron-loss equivalent-circuit objective and analytically eliminates the torque equality constraint, reducing the LMC problem to a one-dimensional scalar minimization problem of the d-axis current. This paper demonstrates that this reduced objective is strictly convex over the admissible scalar domain, allowing for an exact benchmark solution via a scalar convex solver. The proposed method constructs a resistance-aware initial reference from explicit MTPA and MTPV approximations and then applies a one-step scalar Newton-type residual correction using the gradient and Hessian of the reduced loss objective. The initial reference reproduces the known exact surface-mounted PMSM (SPMSM) LMC solution in the SPMSM limit. The correction direction is proved to be a strict descent direction, and a safeguarded step-size ensures loss reduction compared to the initial approximation. The main novelty is the use of the scalar LMC optimality residual as a fixed-cost correction layer for explicit LMC references. The numerical validation is interpreted as steady-state current-reference mapping accuracy and model-based controllable loss-objective verification against the exact scalar optimum, not as hardware drive-efficiency validation. Tests including the no-MTPV case show reduced current-reference and loss-objective gaps while retaining a fixed-cost structure that may support future embedded implementation after validation; hardware transient and efficiency validation under inverter nonlinearity, temperature variation, saturation, and PWM effects remains for future work. Full article

Lightweight, privacy-aware and adversarial robust intrusion detection is required for the proliferation of Internet of Things (IoT) devices. In the Industrial Internet of Things (IIoT), centralized detectors can be compromised by adversarial perturbations via gradient-based attacks, making them susceptible to raw traffic. We suggest Federated Learning-Adaptive Gated Recurrent Unit (FL-AdGRU), a Federated approach that combines a lightweight Gated Recurrent Unit (GRU) classifier with alternating adversarial fine-tuning on each client using FGSM and PGD, without any communication overhead. A two-stage resampling scheme (UCAS-SMOTE) reduces the class-imbalance ratio from 4081:1 to ≈4:1, followed by 61 features being reduced to 40 by a mutual-information selector (MI-SelectK). Under this scenario, FL-AdGRU achieves 99.9% accuracy and 0.999 weighted F1 (+6.5 p.p. over the federated DNN baseline), with no loss of accuracy when facing clean attacks, and boosts Fast Gradient Sign Method FGSM/Projected Gradient Descent (PGD) robustness by +19.3/+19.0 p.p. at the same level of $ϵ$ = 0.1, thus effectively balancing the accuracy–robustness trade-off. It is robust (97.8%/84.2% on UNSW-NB15) and generalizes well to UNSW-NB15, while decaying slowly in skeptical scenarios (≈99.9% weighted F1 for moderate skew, 93.9%/86.7% for severe). Assuring data-locality privacy through exchange of only model weights; defenses against inference attack are left for future work. FL-AdGRU, with a total communication of 43.8 MB (≈50× less than centralized training), is deployable on bandwidth-constrained IIoT networks. Full article

Bridge health monitoring and machine learning are increasingly intertwined for civil engineers and artificial intelligence experts. Bridges’ poor health can result in severe outcomes if not addressed in time. Therefore, continuous monitoring is required to detect any anomaly or damage. Sensors, such as accelerometers, inclinometers, thermistors, etc., can help actively monitor these bridges. The signals from these sensors help record physiological activities. Such activities are helpful for anomaly detection, damage localization, and bridge health predictions with the help of machine learning algorithms. The proposed method extracts features from the dynamic response of a bridge to ambient excitation. It focuses on processing the signal received from different accelerometers installed on a steel railway bridge to determine the location of the damage and the level of the damage predictions. Initially, features are extracted from time-series data; then, they are fed to a deep neural network after some pre-processing. Normal and augmented data are used with different parameter tuning for results. Original data is also subdivided, and the effect of data slicing on the predictions is investigated. The results show that one-fourth of the slicing of the original data gives the best results for training and testing accuracy with a deep neural network. The results show that the reduced matrix representation, particularly the 40 × 40 feature slicing, improved the classification performance for the predefined bridge scenario classes under the considered experimental settings. For bridge scenario classification, the best reported accuracy was 93.54%, while for damage intensity classification the best reported accuracy was 98.21%. In the DNN-based optimizer comparison, the Adam optimizer achieved higher and more stable performance than Stochastic Gradient Descent (SGD), with test accuracies of 92.3% and 93.7% compared with 75.2% and 86.4%, respectively. It is also observed that the Adam optimizer outperformed Stochastic Gradient Descent (SGD) in terms of both damage localization and damage intensity estimation. Full article

The large-scale integration of converter-interfaced renewable energy sources has significantly reduced power system inertia, posing challenges to frequency stability. Although virtual inertia and primary frequency control can enhance the frequency support capability of renewable energy units, their actual performance often deviates from set values due to dynamic response differences among various energy sources (e.g., energy storage, photovoltaic, and wind power) and coupling between inertia and primary regulation power. Existing evaluation methods fail to accurately decouple these components or account for unit-specific dynamic characteristics, leading to considerable estimation errors. To address these issues, this paper proposes a novel estimation method for the frequency regulation capability of renewable energy stations. First, the dynamic frequency response characteristics of synchronous and renewable generators are compared. Then, a decoupling method is developed to separate virtual inertia power from primary frequency regulation power by leveraging their distinct response features. A first-order plus delay time (FOPDT) model is employed to characterize the external frequency response of different renewable energy units. The primary frequency regulation coefficient is estimated using a sliding window integration method, and the virtual inertia time constant is identified via a gradient descent algorithm based on the decoupled inertia power. A hardware-in-the-loop experimental platform is constructed using a real-time digital simulator (RTDS) and phasor measurement units (PMUs) to validate the proposed method. Simulation results show that the estimation errors for energy storage, photovoltaic, and wind power units are 0.63%, 6.38%, and 8.38% for the virtual inertia time constant and 0.45%, 0.72%, and 3.81% for the primary frequency regulation coefficient, respectively. Field test data further confirm the practical applicability and accuracy of the approach. The proposed method enables precise frequency control capability estimation, providing a reliable basis for parameter setting and capacity configuration of frequency regulation resources in low-inertia power systems. Full article

Agricultural pest infestations cause substantial global crop losses. Morphological similarities across species and structural variations across developmental stages render accurate identification a persistently expert-dependent and time-consuming process. Recent deep learning approaches have advanced automated pest classification; however, most efforts have concentrated on architectural design, while optimization strategies have received comparatively little attention. This study proposes a novel optimization framework, hereafter referred to as Adaptive Fractional Gradient Descent (AFGD), that integrates the Grünwald–Letnikov (GL) fractional derivative into the backpropagation process of deep convolutional neural networks. Unlike standard gradient descent, the proposed method maintains a weighted history of past gradients. It dynamically adjusts the fractional order α via Bayesian optimization at regular training intervals, enabling the model to adaptively balance exploiting gradient memory against exploring new gradients throughout training. Experiments conducted on the IP102 benchmark dataset using DenseNet121, ResNet101, and EfficientNetB0 backbones demonstrated consistent accuracy improvements over standard gradient descent across all configurations. In the untrained setting, absolute test accuracy improved by 20.73, 11.51, and 11.01 percentage points for DenseNet121, ResNet101, and EfficientNetB0, although the absolute accuracy levels in this configuration remain modest. Under ImageNet pre-training, the proposed method yielded absolute gains of 6.69, 7.39, and 3.76 percentage points over the corresponding standard gradient baselines, with the highest absolute test accuracy of 70.81% recorded for DenseNet121. These findings indicate that fractional-order gradient control is a promising, architecturally complementary optimization strategy for robust pest classification, with broader implications for deep learning applications in precision agriculture. Full article

Audio deepfake and vocoder fingerprint detectors are increasingly used to identify synthetic speech and attribute it to its generating model. However, their robustness against adversarial perturbations remains unclear across attack algorithms, perturbation domains, detector representations, and vocoder types. This paper presents a focused, quality-aware evaluation of four representative adversarial attacks, namely the Fast Gradient Sign Method (FGSM), Basic Iterative Method (BIM), Projected Gradient Descent (PGD), and Carlini–Wagner (CW) attack, against audio deepfake and vocoder fingerprint detectors. Each attack is implemented in both the waveform domain and the short-time Fourier transform (STFT) magnitude domain. All attacks are optimized against Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networks (AASIST) under a targeted fake-to-real objective and are evaluated on synthetic speech generated by HiFi-GAN, Fullband MelGAN, StyleMelGAN, and Parallel WaveGAN. Attack performance is first measured on the source AASIST detector, after which black-box transferability is assessed on three target detector families: ResNet with Linear Frequency Cepstral Coefficient (LFCC) features, LCNN with Constant-Q Cepstral Coefficient (CQCC) features, and a bidirectional long short-term memory (BiLSTM) detector. The results show that adversarial effectiveness depends strongly on perturbation domain and detector representation. STFT-magnitude PGD transfers strongly to LFCC-based ResNet detectors but has limited effect on CQCC-based and recurrent detectors. In contrast, waveform-domain attacks produce broader transferability across feature-based detectors, with different attacks showing distinct ASR–quality trade-offs. Under the chosen waveform-domain budget, FGSM and BIM preserve transcription-level intelligibility while retaining meaningful black-box transferability, whereas CW provides the strongest overall source-detector and black-box attack performance. To distinguish effective adversarial perturbations from destructive signal degradation, we evaluate audio quality and intelligibility using word error rate (WER) and signal-to-noise ratio (SNR). Overall, the findings show that robustness claims in audio deepfake and vocoder fingerprint detection are limited when adversarial perturbations, black-box transferability, and audio quality are jointly considered. Full article

Continuous monitoring of lower-limb kinematics in natural environments is essential for gait analysis and rehabilitation but remains challenging due to the limitations of optical systems and the inaccuracy of sparse inertial sensor methods. To address this, we propose a high-precision, minimalist wearable system utilizing only three inertial measurement units placed on the pelvis and shanks. In the data preprocessing stage, engineering modifications are made based on the traditional gradient descent algorithm to implement adaptive channel adjustment on the acceleration and magnetic data of a single IMU, aiming to alleviate the impact of motion acceleration and external magnetic interference on the temporal feature manifold. Subsequently, a pure Transformer neural network is utilized to capture long-range temporal dependencies, reconstructing full lower-limb kinematics without relying on rigid biomechanical assumptions. The model was optimized and deployed on an STM32N647 microcontroller to achieve real-time edge inference with a low latency of approximately 17 ms. Experimental results demonstrate that the proposed method achieves a mean absolute error of 2.41° for level walking, significantly outperforming traditional constrained Kalman filter approaches. Furthermore, it maintains high tracking robustness during complex nonlinear movements such as squatting and lunging. In conclusion, this edge-computing-enabled framework provides an accurate, comfortable, and real-time solution for unconstrained human motion capture in daily scenarios. Full article

This paper investigates the inherent conflict between individual and collective interests within multi-coalition games. Unlike traditional noncooperative frameworks where players solely optimize a collective objective, our model incorporates individual player preferences, naturally formulating a constrained multi-objective game. To address this, we introduce an endogenous preference weight factor to scalarize the multi-objective problem into a tractable single-objective game. Furthermore, we propose distributed game strategies equipped with predefined-time convergence to compute the Nash equilibrium of the multi-coalition game, alongside the weight factors. Subsequently, consensus protocols and gradient descent methods are synthesized to locate the weighted Nash equilibrium of the scalarized game. The rigorous predefined-time convergence of the proposed algorithm is established via Lyapunov stability theory. Finally, numerical simulations validate the efficacy and superiority of the proposed framework. Full article

This study addresses distributed decision-making in multi-agent systems under shared constraints. Existing methods often fail to guarantee strict constraint satisfaction or require sensitive parameter tuning. We propose a novel algorithm that integrates a revision protocol with a consensus mechanism. The key innovation is a built-in, parameter-free constraint-checking function within the revision protocol, which automatically halts infeasible strategy updates. This approach enables agents using only local neighbor communication to seek a Generalized Nash Equilibrium (GNE). Theoretical analysis proves that the algorithm converges exponentially. Extensive simulations demonstrate its superiority: it achieves faster convergence and ensures strict per-iteration constraint satisfaction, significantly outperforming traditional gradient descent and penalty-based methods across various network topologies. Full article

We propose the higher-order quasi-Newton (HOQN) method, a hybrid algorithm for unconstrained optimization that combines Newtonian predictors with higher-order correctors derived from vector extensions of the Traub, Chun, and Ostrowski methods, along with quasi-Newton updates of the inverse Hessian using Broyden–Fletcher–Goldfarb–Shanno (BFGS) or Davidon–Fletcher–Powell (DFP) formulas. We demonstrate that the resulting scheme achieves cubic local convergence order, representing a substantial improvement over the superlinear convergence typical of classical quasi-Newton methods, while maintaining a cost of $O\left(n^2\right)$ per iteration. We also analyze variants that incorporate two successive quasi-Newton updates, and show that they retain the same cubic order. Numerical experiments with the benchmark functions of Himmelblau and Freudenstein–Roth confirm the theoretical convergence order and show that the hybrid variants consistently require fewer iterations than BFGS, DFP, and Symmetric Rank-One (SR1). In the case of the Booth function, given its strictly convex quadratic structure, the proposed hybrid methods reach the global minimum in just two iterations and exhibit numerical accuracy superior to that of classical quasi-Newton methods. In addition, limited-memory variants (L-HOQN) are introduced; these are evaluated during the training of a convolutional neural network on the MNIST dataset, where they achieve test accuracies exceeding $99\%$ and outperform L-BFGS and standard stochastic gradient descent (SGD) at all tested learning rates. Full article

Performance of fine motor tasks during the initial phase of space missions is often compromised by the adaptation to microgravity. Since traditional Earth-based training methods are limited and struggle to replicate these conditions without strict time constraints, we propose the training of fine motor tasks with simulated microgravity on earth using an upper limb active exoskeleton. With a model-based control approach, we create a state of microgravity for both arms. To enable realistic microgravity simulation, a suitable model of the human arm is needed. We developed a method to identify the parameters of an arm model by leveraging the computational graph of the inverse dynamics algorithm and utilizing gradient descent to minimize the discrepancy between model and reality. Preliminary data from parabolic flights show that subjects trained with our exoskeleton achieved higher accuracy in a fine motor task during their first exposure to real microgravity compared to untrained subjects. Full article

In underwater acoustic (UWA) orthogonal frequency division multiplexing (OFDM) systems, the orthogonality among subcarriers is highly susceptible to Doppler-induced scaling, leading to severe inter-carrier interference (ICI). This paper proposes a coarse-to-fine Doppler estimation approach for coded orthogonal frequency division multiplexing (OFDM) systems operating in underwater acoustic (UWA) channels. The proposed method first employs the fractional Fourier transform (FRFT) to obtain an initial Doppler factor estimate from a linear frequency modulation (LFM) probe, exploiting the energy concentration property of chirp signals in the fractional domain. This coarse estimate then guides a refinement stage that leverages the cyclic prefix (CP) inherent to each OFDM symbol, enabling symbol-by-symbol Doppler tracking without waiting for the entire packet. As a result, the required memory and processing latency are substantially lower than with full-packet resampling or iterative gradient-descent alternatives. Numerical simulations conducted under both time-invariant and time-variant Doppler conditions demonstrate that the proposed scheme achieves a mean squared error (MSE) below 0.5% at signal-to-noise ratios (SNR) of 5 dB and above. Moreover, the bit error rate (BER) remains within 0.2 dB of an ideal Doppler-free system at a BER of 10⁻³. The combination of low storage demand, symbol-level operation, and robust performance makes the proposed method well-suited for real-time underwater acoustic communication. Full article

Accurate and robust control of robotic joints is essential for high-performance robotic systems. However, conventional proportional–integral–derivative (PID) controllers suffer from limited adaptability when applied to brushless direct current (BLDC) motor-driven joints operating under nonlinear and time-varying conditions. To address this issue, this paper proposes a Radial Basis Function (RBF) neural network-enhanced self-tuning PID control strategy. The RBF neural network serves as an online identifier to approximate the nonlinear dynamics of the BLDC motor and to estimate the system Jacobian online. Based on the estimated Jacobian, the PID gains (K_p, K_i, and K_d) are adaptively updated using a gradient descent mechanism, enabling continuous adjustment to varying operating conditions. Simulation and experimental results demonstrate that the proposed method achieves negligible overshoot, faster settling performance, and improved steady-state accuracy compared with conventional PID and PI controllers. In addition, the proposed controller exhibits enhanced disturbance rejection capability and robust performance under abrupt speed variations and start–stop conditions. The proposed approach effectively combines the simplicity of PID control with the adaptability of neural networks, providing a practical and efficient solution for high-precision robotic joint control. Full article