High Convergence Research Articles

Medical Image Classification Using DL-based Feature Extraction in IoMT

Aim:: Recent advances in Artificial Intelligence (AI) and the addition of Deep Learning (DL) have made it possible to analyse both real-time and historical data from the Internet of Things (IoT). Recently, IoT technology has been implemented in healthcare schemes as IoMT to aid in medical diagnoses. Medical image classification is useful for predicting and identifying serious diseases at an early stage, which is crucial in the diagnostic process. Background:: When it comes to managing, treating, and preventing illness, medical photographs are an essential element of a patient’s health record. However, it is a difficult issue in computer-based diagnostics to classify images using efficient characteristics. Objective:: The study aimed to develop a deep learning-based classification model for feature extraction. Methods:: Levy flight optimization is employed to pick the weight for the classification model optimally. At the end of the day, the optimal weight led to a better classification result and a higher degree of precision when analyzing medical photos for disease. Result:: We tested the proposed results in MATLAB and compared them with conventional methods of classification. The suggested model’s best results include 97.71% accuracy on a brain dataset and 97.2% accuracy on an Alzheimer’s disease dataset. Conclusion:: The proposed algorithm’s high rate of convergence proves that it can successfully balance the exploration and exploitation phases by avoiding capturing in local optimization and classifying thresholds rapidly. In light of the need for improved accuracy, precision, and computational speed in clinical picture classification, a novel approach based on soft sets has been presented.

Making plaque assessment easier - a validation study of simplified versions of the Marginal Plaque Index.

The assessment of plaque indices may be time-consuming and error-prone. Simplification of these indices may increase their utility without compromising their validity. The aim of this study was to evaluate the validity of two simplified versions of the Marginal Plaque Index (MPI). Two simplified versions of the MPI as well as the Plaque Control Record (PCR) were derived from full-scale MPI assessments in two studies with four age groups (N = 42 10-year-olds; N = 24 15 year-olds; N = 53 university students (18y-33y); N = 66 parents (32y-57y). Correlations with the Turesky modification of the Quigley-Hein Index (TQHI) and the Papillary Bleeding Index (PBI) were calculated. The indices derived from the MPI showed high convergence with each other (all r ≥ 0.94) and with the TQHI (r ≥ 0.80). The concurrent validity of the MPI with the PBI was equal to that of the TQHI in all age groups. The simplified versions of the MPI and the PCR show a lower convergent validity with the PBI than the MPI within parents (p < 0.05). In the other age groups, their convergent validity was equal to that of the MPI. Simplification of the MPI does not affect its convergent validity with other plaque indices but may reduce its concurrent validity with the PBI in middle-aged adults.

Penalty 4-Node Quadrilateral Element Formulation for Axisymmetric Couple Stress Problems

To address the issue of size effects in axisymmetric deformation of small-scale solids, this work proposes a 4-node 12-DOF element for axisymmetric problems based on the consistent couple stress theory (CCST), following the framework of the unsymmetric finite element method. With the use of the penalty function method, an independently assumed rotational field is introduced into the virtual work principle to approximate the physical rotation, ensuring the satisfaction of the C1 continuity requirement of the CCST in a weak form. As a benefit, the enriched C0 isoparametric-based interpolation is employed to construct the test functions for displacement and rotation. Furthermore, the force-stress field that satisfies the equilibrium equations related to axisymmetric deformation is employed as the element’s force-stress trial function. In order to circumvent locking issues, reduced integration is employed in the penalty stiffness integration process. The numerical results demonstrate that the new element exhibits high computational accuracy and convergence rate in both static and modal analysis problems, effectively capturing size-dependent phenomena.

A Lightweight and Efficient Multimodal Feature Fusion Network for Bearing Fault Diagnosis in Industrial Applications

To address the issues of single-structured feature input channels, insufficient feature learning capabilities in noisy environments, and large model parameter sizes in intelligent diagnostic models for mechanical equipment, a lightweight and efficient multimodal feature fusion convolutional neural network (LEMFN) method is proposed. Compared with existing models, LEMFN captures rich fault features at multiple scales by combining time-domain and frequency-domain signals, thereby enhancing the model’s robustness to noise and improving data adaptability under varying operating conditions. Additionally, the convolutional block attention module (CBAM) and random overlapping sampling technology (ROST) are introduced, and through a feature fusion strategy, the accurate diagnosis of mechanical equipment faults is achieved. Experimental results demonstrate that the proposed method not only possesses high diagnostic accuracy and rapid convergence but also exhibits strong robustness in noisy environments. Finally, a graphical user interface (GUI)-based mechanical equipment fault detection system was developed to promote the practical application of intelligent fault diagnosis in mechanical equipment.

Marx-Biased Technical Change and International Profit-Rate Convergence: The Case of Mexican Manufacturing

The article studies the fall in the output/capital (OC) ratio in Mexican manufacturing, contrasting the hypotheses of high labor costs versus convergence toward the United States as drivers of Marx-biased technical change. While econometric estimations reject the labor-cost hypothesis—as do the evidence from wage-profit curves—they support the convergence one, showing in addition that convergence increased with the degree of US-Mexican industrial integration. Downward OC convergence, moreover, led to downward convergence in profit rates. Convergence aside, capital accumulation itself pushed the OC ratio down, confirming a tendency of accumulation to depress the profit rate. The estimations—for the change in the profit rate and the OC ratio—rely on a comprehensive annual panel of 17 Mexican manufacturing industries over the period 1992–2019. JEL Classification: C23, F15, F63, O14, O33

Two-Step Fifth-Order Efficient Jacobian-Free Iterative Method for Solving Nonlinear Systems

This article introduces a novel two-step fifth-order Jacobian-free iterative method aimed at efficiently solving systems of nonlinear equations. The method leverages the benefits of Jacobian-free approaches, utilizing divided differences to circumvent the computationally intensive calculation of Jacobian matrices. This adaptation significantly reduces computational overhead and simplifies the implementation process while maintaining high convergence rates. We demonstrate that this method achieves fifth-order convergence under specific parameter settings, with broad applicability across various types of nonlinear systems. The effectiveness of the proposed method is validated through a series of numerical experiments that confirm its superior performance in terms of accuracy and computational efficiency compared to existing methods.

High-efficiency MPPT strategy for PV Systems: Ripple-free precision with comprehensive simulation and experimental validation

This paper presents a newly developed maximum power point (MPP) tracking algorithm (MPPT) to boost the tracking performance of solar photovoltaic (PV) systems. By functioning PV arrays at their MPP and eliminating the ripple problem in the converter's output, the newly developed strategy can markedly improve the precision of tracking and overcome the issues faced by many existing algorithms. The proposed strategy uses an MPP voltage boundary to control the PV voltage and directly generate the required duty cycle using a mathematic expression, resulting in a high convergence speed, drift problem avoidance, and zero MPP fluctuations. To prove and validate the robustness of tracking of the suggested MPPT strategy, both simulation and experiment validations based on the MATLAB/Simulink and dSPACE DS1104 board platforms, respectively, are carried out under swift variations in solar irradiance. The effectiveness of the proposed approach is evaluated by comparing it to the InC algorithm in terms of tracking accuracy. The results from both simulations and experiments demonstrate that the suggested strategy surpasses the InC approach in multiple aspects. It exhibits a shorter response time, eliminates the ripple problem, and achieves a superior tracking efficiency of 99.60 %.

Innovative hybrid energy storage systems with sustainable integration of green hydrogen and energy management solutions for standalone PV microgrids based on reduced fractional gradient descent algorithm

This paper investigates innovative solutions to enhance the performance and lifespan of standalone photovoltaic (PV)-based microgrids, with a particular emphasis on off-grid communities. A major challenge in these systems is the limited lifespan of batteries. To overcome this issue, researchers have created hybrid energy storage systems (HESS) along with advanced power management strategies. This study introduces innovative multi-level HESS approaches and a related energy management strategy designed to alleviate the charge/discharge stress on batteries. Comprehensive Matlab Simulink models of various HESS topologies within standalone PV microgrids are utilized to evaluate system performance under diverse weather conditions and load profiles for rural site. The findings reveal that the proposed HESS significantly extends battery life expectancy compared to existing solutions. Furthermore, the paper presents a novel energy management strategy based on the Reduced Fractional Gradient Descent (RFGD) algorithm optimization, tailored for hybrid systems that include photovoltaic, fuel cell, battery, and supercapacitor components. This strategy aims to minimize hydrogen consumption of Fuel Cells (FCs), thereby supporting the production of green ammonia for local industrial use. The RFGD algorithm is selected for its minimal user-defined parameters and high convergence efficiency. The proposed method is compared with other algorithms, such as the Lyrebird Optimization Algorithm (LOA) and Osprey Optimization Algorithm (OOA). The RFGD algorithm exhibits superior accuracy in optimizing energy management, achieving a 15 % reduction in hydrogen consumption. Its efficiency is evident from the reduced computational time compared to conventional algorithms. Although minor losses in computational resources were observed, they were substantially lower than those associated with traditional optimization techniques. Overall, the RFGD algorithm offers a robust and efficient solution for enhancing the performance of hybrid energy systems.

Exploring Normalization for High Convergence on Federated Learning for Drones

The usage of mobile devices like drones has been increasing in various fields, ranging from package delivery to emergency services and environmental monitoring. Intelligent services increasingly use the processing power of these devices in conjunction with techniques such as Federated Learning (FL), which allows machine learning to be carried out in a decentralized way using data accessed by clients or devices. However, in normal operations, the data accessed by clients is distributed heterogeneously among themselves, negatively impacting learning results. This article discusses the normalization in Federated Learning local training to mitigate results obtained in heterogeneous distributions. In this context, we propose Federated Learning with Weight Standardization on Convolutional Neural Networks (FedWS) and evaluate it with Batch Normalization, Layer Normalization, and Group Normalization in experiments with heterogeneous data distributions. The experiments demonstrated that FedWS achieved higher accuracy results ranging from 3% to 6% and reduced the computational and communication costs between 25% and 40%, being more suitable for use in devices with computational resource limitations.

A Harmonic Response Topology Optimization Method Based on Hybrid Cellular Automata

Abstract Hybrid cellular automaton (HCA) method is a topology optimization method with high convergence efficiency. Harmonic response topology optimization problem has been widely researched and common frequency-domain methods have been applied to solve this problem. Despite this, little work has so far been undertaken to utilize HCA method to improve the calculation efficiency of harmonic response topology optimization. In this paper we present a methodology which applies HCA method to solve harmonic response topology optimization problem. Mode displacement method (MDM), mode acceleration method (MAM) and full method (FM) are common frequency-domain methods which are respectively utilized as harmonic response analysis methods in this study. The examples under rotating loads demonstrate that the feasibility of this methodology at one specific frequency excitation and multiple frequencies excitation. This paper presents and implements a high convergence efficiency method for harmonic response topology optimization, this method can converge in 20–30 iterations. This method can extend the application range of HCA method and improve the optimization efficiency of harmonic response topology optimization.

Achieving High Convergence Rates by Quasi-Monte Carlo and Importance Sampling for Unbounded Integrands

Achieving High Convergence Rates by Quasi-Monte Carlo and Importance Sampling for Unbounded Integrands

State of Charge Estimation for Lithium-Ion Battery Based on Fractional-Order Kreisselmeier-Type Adaptive Observer

For the safe and efficient use of lithium-ion batteries, the state of charge (SOC) is a particularly important state variable. In this paper, we propose a method for the online estimation of SOC and model parameters based on a fractional-order equivalent circuit model. Firstly, we constructed a fractional-order battery model that includes pseudo-capacitance and determined the values of the circuit elements offline using the least squares method from actual input–output data based on the driving profile of an automobile. Compared to the integer-order battery model, we confirmed that the proposed fractional-order battery model has higher accuracy. Secondly, we constructed a fractional-order Kreisselmeier-type adaptive observer as an observer that performs state estimation and parameter adjustment simultaneously. Applying the general adaptive law to the battery model results in a redundant design with many adjustable parameters, so we proposed an adaptive law that reduces the number of adjustable parameters without compromising the stability of the observer. The effectiveness of the proposed method was verified through numerical simulations. As a result, the high estimation accuracy and convergence of the proposed adaptive law were confirmed.

Q-learning improved golden jackal optimization algorithm and its application to reliability optimization of hydraulic system

To endow the prey with intelligent movement behavior and improve the performance of Golden Jackal Optimization (GJO), a Q-learning Improved Gold Jackal Optimization (QIGJO) algorithm is proposed. This paper introduces five update mechanisms and proposes double-population Q-learning collaborative mechanism to select appropriate update mechanisms to improve GJO performance. Additionally, a new convergence factor is incorporated to enhance convergence capability of GJO. QIGJO demonstrates excellent performance across 23 benchmark functions, CEC2022, and three classical engineering design problems, indicating high convergence accuracy and significantly enhanced global exploration capability. The reliability optimization model of the hydraulic system for concrete pump trucks was established based on a Continuous-time Multi-dimensional T-S dynamic Fault Tree (CM-TSdFT), considering the two-dimensional factors of operating time and number of impacts. Utilizing QIGJO to optimize this model yielded excellent results, providing valuable methodological support for reliability optimization of hydraulic systems.

Improving the efficiency of a non-ideal grid coupled to a photovoltaic system with a shunt active power filter using a self-tuning filter and a predictive current controller

Introduction. Recently, photovoltaic (PV) systems are increasingly favored for converting solar energy into electricity. PV power systems have successfully evolved from small, standalone installations to large-scale, grid-connected systems. When the nonlinear loads are connected to a grid-tied PV system, the power quality can deteriorate due to the active power supplied by the PV array, there’s a noticeable decline in the quality of power delivered to consumers. Its combination with the shunt active power filter (SAPF) enhances system efficiency. Consequently, this integrated system is adept at not only powering local loads but also at compensating for reactive power and filtering out harmonic currents from the main grid. The novelty of the work describes how an operation of a small scale PV system connected to the low voltage distribution system, and nonlinear load can be achieved, the investigation aims to analyze the system’s behavior and elucidate the advantages of employing various control algorithms. These proposed algorithms are designed to ensure a unity power factor for the utility grid while prioritizing high convergence speed and robustness against load power fluctuations across different levels of solar irradiation affecting the PV modules. The purpose of this work is to enhance the dynamic performance of the SAPF by cooperatively using a self-tuning filter (STF) based instantaneous active and reactive power method (PQ) with a novel predictive current control, enhance the system resilience, ensure optimal management of the total active power between the PV system, the electrical network and the non-linear load by integrating the functionalities of the SAPF under different levels of solar irradiation and maintain the DC-link capacitor voltage constant. Methods. A novel predictive current controller is designed to generate the switching signals piloted the three phase source voltage inverter, also a novel algorithm of instantaneous active and reactive power is developed, based on STF, to extract accurately the harmonic reference under non ideal grid voltage, also the perturb and observe algorithm is used to extract, under step change of solar irradiation, the maximum power point tracking of the PV module and the PI controller is used to maintain constant the DC-link capacitor voltage of the SAPF. Results. The efficacy of the proposed system is primarily centered on the grid side, and the performance evaluation of the control system is conducted using the STF based PQ algorithm and predictive current control. In addition, comprehensive testing encompasses all modes of operation, including scenarios involving distorted voltage sources, step changes in solar radiation, and variations in nonlinear loads. Results highlight superior performance in both transient and stable states, affirming the robustness and effectiveness of the proposed controllers. Practical value. The total harmonic distortion value of the grid current for all tests respects the IEEE Standard 519-1992. References 21, tables 7, figures 25.

Gold Self-Relay Catalysis Enabling Regioselective Bicyclization toward [5]Azahelicenes.

A gold self-relay catalysis driving a double annulation cascade starting from soft electron-biased 1,2-di(o-aminoaryl)alkynes and aldehydes is reported, enabling regioselective access to produce a series of [5]azahelicenes depending on the substitution pattern in generally good yields under mild conditions. This protocol exploits and unifies the π- and σ-Lewis acid capability of gold catalysts, featuring high molecular convergence, broad substrate flexibility, and good functional group compatibility and regioselectivity.

Research on key aircraft identification method based on time-effective network

For the problem of identifying key aircraft in air traffic situation, existing research has failed to fully consider the spatiotemporal effects in the actual operation of air traffic. Therefore, a key aircraft identification method based on time-effective network is proposed. The convergence relationship and complexity between aircraft are used to construct a time-effective network model through the neighbor topology overlap coefficient, and the key aircraft is determined based on the eigenvector centrality. A network attack is carried out on the key aircraft nodes to observe the changes in sector complexity, and compared with the attack based on static network indicators, an improved genetic algorithm is used to assign new entry time to the aircraft nodes deleted by the network attack, so as to verify the selection effect of key aircraft. The actual data verification shows that compared with the static network attack, this method is more efficient in reducing the average complexity of the sector when removing key aircraft. The improved genetic algorithm has higher convergence when solving the key aircraft entry time allocation problem, making the sector complexity more stable within a certain period of time. Analysis of the control effect of key aircraft shows that the time-effective network method can more accurately identify aircraft that have a greater impact on sector complexity within a period of time than the static network.

A Morley Type Triangular Finite Element with High Convergence for the Biharmonic Problem

In this work, we construct a theoretical framework to develop non C0 Morley type nonconforming high-convergence elements for biharmonic problems. For each element domain, P3 should be included in the space of shape functions. Besides the degrees of freedom of Morley elements, we add the integrals and first-order moments of the normal derivatives on edges. The choice of degrees of freedom and shape function space guarantees the possibility of improving the convergence order. As an application, we specifically construct a Morley type element on triangular meshes. Lastly, numerical experiments are carried out to verify the feasibility of the element.

A Relaxed Tseng’s Method-based Algorithm for Solving Split Variational Inequalities with Multiple Output Sets

In this paper, the split variational inequality problem with multiple output sets is considered in a more general framework of reflexive Banach spaces. We introduce a relaxed Tseng extragradient method for approximating the solution of this problem when the cost operators are pseudomonotone and non-Lipschitz. Moreover, our proposed algorithm employs the inertial technique and a new self-adaptive step size to guarantee high rate of convergence. Strong convergence of the algorithm is proved in this case without the need for the sequential weak continuity condition. Finally, some numerical experiments are given to show the applicability of our proposed algorithm.

Dynamic evolution of competing same-dip double subduction: New perspectives of the Neo-Tethyan plate tectonics

Same-dip double-subduction (SDDS) systems are widely reported from present as well as past complex convergent plate tectonic configurations. However, the dynamics of their evolution is poorly understood, which is crucial to conceptualize anomalous subducting slab kinematics and associated observed geological phenomena, such as irregular trench migration rates, high convergence velocities, and slab break-off. To bridge this gap, we develop dynamic thermo-mechanical subduction models and investigate the initiation and evolution of SDDS systems, considering three different initial plate configurations: oceanic, oceanic-continental and multiple continental settings, based on Neo-Tethyan paleo-reconstructions. Each model offers new insights into the complex tectonic history of the major Neo-Tethyan subduction zones, particularly the Indo–Eurasian and Andaman convergent systems. We evaluate the slab-slab interactions, trench and subduction kinematics, inter-plate reorganization, and temporally varying mantle flow patterns involved in the dynamic evolution of these SDDS systems. The oceanic SDDS model simulations reveal that a sizable oceanic plate can initiate two subduction zones synchronously, and they evolve unequally in a competing mode, leading to exceptionally high convergence rates (∼16–17 cm/year) for a prolonged duration (∼7–8 Myr). This finding explains the coeval activity of coupled subduction zones in the Indo-Eurasia convergence during the Cretaceous evolution of the Neo-Tethys. We further implement a corresponding single subduction model to assess the additional effects of competing slab kinematics in an oceanic SDDS setting. The ocean-continent SDDS model, on the other hand, localizes subduction preferentially at passive margins between the oceanic plate and the continental block, forming double subduction zones that grow almost equally to form a spreading centre between the two trenches. These model results allow to reconstruct the Cenozoic evolution of the eastern Neo-Tethyan region, which ultimately led to the development of the Andaman subduction zone. We also show the post-Cretaceous evolution of the Indo–Eurasian collision zone as a consequence of the SDDS dynamics in presence of multiple continental blocks. These dynamics facilitated slab break-off, transforming the SDDS into a single subduction system in a relatively short time frame (∼3 Myr). We finish with a synthesis of the paleo-reconstructions of the Neo-Tethys in the perspective of these SDDS models.

Assessment of the comparability of human facial dimensions obtained using direct anthropological measurements and photographs

Introduction. This research examines and evaluates the comparability of a specific set of anthropometric measurements of facial features obtained through direct measurements and frontal portrait photographs. With the growing trend towards utilizing photography instead of direct measurements, the question of the convergence and comparability between these two methods has become increasingly significant. Materials and methods. The study was conducted on a sample population of Chuvash individuals, who represent the indigenous population in the Chuvash Republic of the Russian Federation. The material for the study comprised cephalometric data and anthropological photographs of individuals from the Chuvash Republic, surveyed during the 2002 expedition under the guidance of V.A. Batsevich. The sample size consisted of 72 individuals, with 35 women and 37 men. Prior to statistical analysis, the male and female participants were divided into two age groups to account for potential age-related variations. Statistical analysis was conducted using the SPSS software package, applying Cronbach's alpha method. Results. In this study, gender differences were identified for a range of anthropometric indicators, demonstrating high convergence when employing two facial measurement methods. The obtained values of dimensions were generally comparable in terms of reliability and consistency. Conclusion. The research has revealed that the method of obtaining linear facial dimensions from portraits can serve as an analog/alternative to direct anthropometric measurements. In men, latitudinal dimensions are better correlated with each other, whereas in women they are high-altitude. The obtained results align well with existing data on this topic, highlighting the potential viability of facial portrait-based measurements in anthropometric studies.