To address the challenge of detecting internal defects in medium-thick titanium alloy laser welds, a combined simulation and experimental study on ultrasonic testing was conducted. A finite element model employing a 5 MHz shear wave angle transducer for inspecting titanium alloy welds was established. An ultrasonic testing system was developed, incorporating a DPR300 pulser-receiver (JSR Ultrasonics, Pittsford, NY, USA) and an MSO5204 oscilloscope (RIGOL, Suzhou, China), and was calibrated using standard reference blocks. The inspection results for four prefabricated internal defects at various depths demonstrated that all defects were effectively detected, with the minimum detectable equivalent defect size reaching 1 mm. The measured signal-to-noise ratio (SNR) averaged 17.6 dB, validating the high sensitivity of the proposed system. The mean absolute error for defect localization was 0.438 mm, achieving a positioning accuracy better than 0.5 mm. This study indicates that the pro-posed method enables effective detection and accurate localization of internal defects in titanium alloy laser welds, providing critical technical support for laser welding quality assessment.

A method based on microwave-induced combustion (MIC) was applied for medicinal plants digestion allowing further chlorine determination by potentiometry using ion-selective electrode (ISE). Sample masses ranging from 500 to 1000 mg were evaluated for MIC digestion. Water and 10, 25, 50, and 100 mmol/L NH4OH were investigated as absorbing solutions. The accuracy of the proposed method was evaluated by using certified reference materials (CRMs), by recovery tests (500 µg/g), and also by comparison with the results obtained by inductively coupled plasma optical emission spectrometry (ICP-OES) after microwave-assisted alkaline extraction (MAE). Using water or NH4OH solutions (10 to 100 mmol/L), recoveries close to 100% and relative standard deviation lower than 5% were obtained. Results were in agreement with CRMs values (better than 95%) and also with those values obtained by using the MAE method. The main advantage of the proposed method was the complete combustion of high sample mass (1000 mg) resulting in low quantification limit (12.5 µg/g) and chlorine determination at low concentration by ISE. Another advantage of the proposed method was the high chlorine stability in digests (up to 30 days of storage) even using water as absorbing solution, which is in agreement with green analytical chemistry recommendations. Finally, the proposed MIC method was applied for commercial medicinal plants and the chlorine concentration was in the range of 59.4 ± 1.4 to 2038 ± 70 µg/g. The proposed MIC method was considered suitable for quality control for chlorine determination in medicinal plants.

Cyberattacks targeting critical infrastructure such as the power grid pose significant risks to the reliability and security of energy networks. Among the cyberattacks, false data injection (FDI) attacks, which manipulate power grid data, is a serious concern due to their potential to cause widespread disruptions and damage. The 2015 Ukraine power grid cyber attack demonstrated its destructive potential and revealed the real-world impact of such threats. It is especially challenging to detect AC FDI attacks since they are based on the AC power flow. Traditional cyberattack detection methods are primarily based on rule-based or statistical anomaly detection. They often fall short in detecting these cyber threats due to their design based on static thresholds. Recent advances in deep learning show promise in improving detection accuracy and adaptability. Building on these advances, this paper proposes a novel physics-informed approach for detecting AC FDI attacks in the power grid, integrating graph attention convolutional networks (GACN), long short-term memory (LSTM) networks, and unscented Kalman filters (UKF). GACN uses the physical topology of the power grid to capture spatial dependencies and correlations between grid components, while LSTM models temporal dynamics. In addition, UKF enhances the detection capabilities of the proposed method as a second layer of protection. Extensive simulation studies using PSCAD/EMTDC datasets demonstrate that the pro posed method outperforms several baseline models in detecting sophisticated cyber threats in the power grid. These baseline models encompass a range of simpler machine learning and deep learning architectures.

The benefits of the Internet of Medical Things (IoMT) in providing seamless healthcare to the world are at the forefront of technological advancement. However, security concerns of any IoMT systems are high since they threaten to compromise personal information of patients and can even cause health hazards. Researchers are exploring the use of various techniques to ensure a high level of security of IoMT systems. One key concern is that the computing power of any Internet of Things (IoT) device is relatively low, hence mechanisms that require low computational power are appropriate for designing Intrusion Detection Systems (IDS). In this research work, a blockchain IDS coalition is proposed for securing IoMT networks and devices. The blockchain ledger is compact and uses less processing resources. Additionally, the ledger requires less communication overhead. The cryptographic hashes in the suggested architecture ensure complete data secrecy and integrity between parties who are trusted and those who are untrustworthy. Peer-to-peer networks in both central and cluster networks are also included in this work for complete decentralization. The proposed model can counter various attacks, including Denial of Service (DoS), anonymity attacks, impersonation attacks, Man-In-The-Middle (MITM), and Cross-Site Scripting (XSS). The proposed method achieved an F1- score as high as 100% and reported an AUC value of over 99%.

Block hybrid linear multistep method was proposed to overcome the Dahl Quist order barrier for linear multistep methods. This research aims to answer questions relating to the convergence, accuracy, and effectiveness of the block hybrid method when utilized to obtain the solution of Initial Value Problems (IVPs). In this research, an order (k+3) block hybrid method applicable to obtain the direct solution of IVP’s of ordinary differential equations (ODEs) is presented. Collocation and interpolation of power series at finely selected grid points were used to improve the method's consistency, convergence, accuracy and zero stability. Linear problems weresolved to show the accuracy and efficiency of the proposed method, and the error obtained from the comparison of exact and approximate results shows that the proposed method is effective in solving the class of problem.

Communication learning is an important research direction in the multiagent reinforcement learning (MARL) domain. Graph neural networks (GNNs) can aggregate the information of neighbor nodes for representation learning. In recent years, several MARL methods leverage GNN to model information interactions between agents to coordinate actions and complete cooperative tasks. However, simply aggregating the information of neighboring agents through GNNs may not extract enough useful information, and the topological relationship information is ignored. To tackle this difficulty, we investigate how to efficiently extract and utilize the rich information of neighbor agents as much as possible in the graph structure, so as to obtain high-quality expressive feature representation to complete the cooperation task. To this end, we present a novel GNN-based MARL method with graphical mutual information (MI) maximization to maximize the correlation between input feature information of neighbor agents and output high-level hidden feature representations. The proposed method extends the traditional idea of MI optimization from graph domain to multiagent system, in which the MI is measured from two aspects: agent features information and agent topological relationships. The proposed method is agnostic to specific MARL methods and can be flexibly integrated with various value function decomposition methods. Considerable experiments on various benchmarks demonstrate that the performance of our proposed method is superior to the existing MARL methods.

Graph convolutional networks (GCNs) have shown great prowess in learning topological relationships among electroencephalogram (EEG) channels for EEG-based emotion recognition. However, most existing GCN-only methods are designed with a single spatial pattern, lacking connectivity enhancement within local functional regions and ignoring the data dependencies of EEG original data. In this article, hierarchical dynamic GCN (HD-GCN) is proposed to explore dynamic multilevel spatial information among EEG channels, with discriminative features of EEG signals as auxiliary information. Specifically, representation learning in topological space consists of two branches: one for extracting global dynamic information and one for exploring augmentation information in local functional regions. In each branch, a layerwise adjacency matrix is utilized to enrich the expressive power of GCN. Furthermore, a data-dependent auxiliary information module (AIM) is developed to capture multidimensional fusion features. Extensive experiments on two public datasets, SJTU emotion EEG dataset (SEED) and DREAMER, demonstrate that the proposed method consistently exceeds state-of-the-art methods. Interpretability analysis of the proposed model is performed, discovering the active brain regions and important electrode pairs related to emotion.

Hyperspectral band selection has become a key focus in hyperspectral image processing as it reduces the spectral redundancy and computational overhead, thereby improving classification performance. However, optimal band selection remains challenging due to its combinatorial nature. Although numerous metaheuristic algorithms have been introduced in recent years to address this problem, achieving an effective balance between exploration and exploitation continues to pose a major challenge. This paper proposes a novel approach that combines a parameter-free binary Jaya algorithm with a mutation operator to enhance exploration and maintain solution diversity within the search space. We employ Opposition-based Learning (OBL) for population initialization and Quasi-Reflection reinitialization strategy to add diversity whenever fitness stagnation occurs. To simultaneously improve classification performance and band reduction we adopt weighted sum multi-objective fitness function that minimizes redundancy and enhances model generalization. Our proposed method is evaluated using three benchmark datasets, namely Indian Pines, Pavia University, and Salinas. Experimental results demonstrate that the pro-posed method outperforms recent metaheuristic-based band selection techniques. Its superior performance makes it well suited for various HSI applications.

Abstract: The proper screw rotor cutting process is essential to obtain a precise rotor profile; however, it is a costly and high hazard if studied in a practical method. This study introduces an analytical screw cutting method to ensure the cutting process is running well and acquire the expected rotor profile. Two distinct cutter types, a cutter with a single curve and multiple inserts cutting edge, were applied. The analytical screw cutting method was developed according to the cutter-workpiece engagement model. The result reveals that the analytical screw cutting model using various cutters can generate identical simulated profiles and close to the original rotor profile. In addition, the virtual machining verification using VERICUT software was conducted to evaluate the proposed method. Conclusively, the analytical screw cutting method is reliable and realistic to be applied in screw rotor milling.

Abnormalities in cardiac function arise irregularly and typically involve multimodal electrical, mechanical vibrations, and acoustics alterations. This article proposes an Electro-Mechano-Acoustic (EMA) activity model for mapping the complete macroscopic cardiac function to refine the systematic interpretation of cardiac multimodal assessment. We abstract this activity pattern and build the mapping system by analyzing the functional comparison of the heart pump and Electronic Fuel Injection (EFI) system from the multimodal characteristics of the heart. Electrocardiogram (ECG), seismocardiogram (SCG) & Ultra-Low Frequency seismocardiogram (ULF-SCG), and Phonocardiogram (PCG) are selected to implement the EMA mapping respectively. First, a novel low-frequency cardiograph compound sensor capable of extracting both SCG and ULF-SCG is proposed, which is integrated with ECG and PCG modules on a single hardware device for portable dynamic acquisition. Afterward, a multimodal signal processing chain further analyses the acquired synchronized signals, and the extracted ULF-SCG is shown to indicate changes in heart volume. In particular, the proposed method based on waveform curvature is used to extract 9 feature points of the SCG signal, and the overall recognition accuracy reaches over 90% in the data collected by EMA portable device. Ultimately, we integrate the portable device and signal processing chains to form the EMA cardiovascular mapping system (EMACMS). As a next-generation system solution for cardiac daily dynamic monitoring, which can map the mechanical coupling and electromechanical coupling process, extract multi-characteristic heart rate variability (HRV), and enable extraction of important time intervals of cardiac activity to assess cardiac function.

Hyperspectral band selection has become a key focus in hyperspectral image processing as it reduces the spectral redundancy and computational overhead, thereby improving classification performance. However, optimal band selection remains challenging due to its combinatorial nature. Although numerous metaheuristic algorithms have been introduced in recent years to address this problem, achieving an effective balance between exploration and exploitation continues to pose a major challenge. This paper proposes a novel approach that combines a parameter-free binary Jaya algorithm with a mutation operator to enhance exploration and maintain solution diversity within the search space. We employ Opposition-based Leaning (OBL) for population initialization and Quasi-Reflection reinitialization strategy to add diversity whenever fitness stagnation occurs. To simultaneously improve classification performance and band reduction we adopt weighted sum multi-objective fitness function that minimizes redundancy and enhances model generalization. Our proposed method is evaluated using three benchmark datasets, namely Indian Pines, Pavia University, and Salinas. Experimental results demonstrate that the pro-posed method outperforms recent metaheuristic-based band selection techniques. Its superior performance makes it well suited for various HSI applications.

With the development of science and technology, more and more images need to be recognized and categorized. Although the classical Bag of Words (BoW) model has played a great role in the past, there are still many limitations about it, i.e. low precision and accuracy, high complexity of computation, etc. In this paper, it is improved and extended from four ways. Firstly, the features filtered from the background are sampled to reduce the influence of background noise. Secondly, the spatial relationship among all features is integrated with the classical BoW vector to improve the accuracy of recognition and categorization. Thirdly, vocabulary tree is constructed by applying hierarchical K mean value, in order to obtain more reasonable vocabulary list and greatly reduce the clustering time. Fourthly, a weighted visual word histogram is considered, in order to stand out the essential difference among images. At last, some experiments are conducted to show the advantage of the proposed method.

Web apps hold important information, such as login tokens and individual data, and cybercriminals repeatedly target attackers. Cross-site scripting is one of the most frequent vulnerabilities in web apps. Several techniques and patents are used to mitigate these vulnerabilities. Several 100 articles from a review of research papers published between 2005 and 2023 were considered. This paper reviewed different techniques and tools to detect cross-site scripting attacks, and it will be helpful to understand, analyze, and develop a strategy to deal with them. This paper focuses on different methods and tools for identifying cross-site scripting (XSS) attacks. Also, it depicts the strengths and shortcomings of the existing proposed method. Additionally, it will help to understand existing open issues or challenges faced by previous researchers.

High-speed trains are susceptible to unexpected events such as strong winds and equipment failures, which can result in deviations from the scheduled timetable. As the density of traffic increases, these delays can quickly spread to other trains, eventually leading to conflicts in the timetable. To ensure the efficiency of high-speed railways, quickly resolving potential conflicts and generating appropriate rescheduling schemes are essential. The existing hierarchical structure of train control and online rescheduling tends to be inefficient in terms of information communication and can even lead to unfeasible rescheduled timetables and trajectories. To address these issues, an integrated structure of timetable rescheduling and train trajectory optimization is proposed by introducing the train minimum running time into the process of timetable rescheduling and using the adjusted running time as the objective of trajectory optimization. The integration model is formulated by considering the constraints of timetable rescheduling such as the maximum number of trains overtaking trains, platforms at stations, and the priority of the train, as well as the constraints of trajectory optimization. A deep reinforcement learning (DRL)-based approach is proposed to solve the problem. Numerical experiments are conducted on a segment of the Beijing-Shanghai high-speed railway line, using adapted data to demonstrate the effectiveness of the proposed method in rescheduling timetables and optimizing train trajectories. The results show that the integrated rescheduled timetable and the optimized train trajectory can be generated simultaneously and the computation time exhibits a linear increase with respect to the size of the problem.

This study reports an optically-induced dielectrophoretic (ODEP) force-based microfluidic platform for live and dead cell separation and collection. ODEP forces are used to separate the live and dead cells due to their opposite responses to an ODEP force. Combining the flow control in a microfluidic system, the live and dead cells can be separated and subsequently collected in an efficient and effective manner. The operating conditions of the ODEP force for manipulating the live and dead chondrocytes is characterized, and separation performance is experimentally evaluated. Results revealed that an applied voltage of 8 V resulted in a maximum difference of manipulation force for the live (49.4 pN) and dead (-20.1 pN) cells. Results of further separation experiments showed that the recovery rate and purity of the isolated live cells was as high as 78.3±6.8 % and 96.4 ±2.2 %, respectively. Overall, the proposed method is found to be particularly valuable for biological research which requires the isolation of highly pure live or dead cells.

This study aims to control indoor temperatures in an air-conditioned room to ensure the occupant’s thermal comfort while minimizing energy consumption. In the literature, controlled simulations of air conditioning systems usually assume that the indoor air is perfectly mixed. This assumption provides little information on spatial temperature and air flow. By contrast, this study deals with imperfectly mixed air. A computational fluid dynamics method is used to model an air-conditioned room and links this model with controllers. A self-tuning controller can monitor plant changes based on recursive estimation and adjusts control parameters to meet desired performance. Therefore, this study develops self-tuning controllers to control room temperature. Disturbances of varying temperature are exerted to investigate control performance. This paper compares the performance of a self-tuning linear quadratic controller and a self-tuning proportional-integral-derivative (PID) controller. Simulation results show that both controllers track desired temperatures well. Compared with the self-tuning PID controller, the self-tuning linear quadratic controller yields less overshoot with a slower response. The proposed method in this study is validated by experimental results.

An adaptive state aggregation Q-Learning method, with the capability of multi-agent cooperation, is proposed to enhance the efficiency of reinforcement learning (RL) and is applied to box-pushing tasks for humanoid robots. First, a decision tree was applied to partition the state space according to temporary differences in reinforcement learning, so that a real valued action domain could be represented by a discrete space. Furthermore, adaptive state Q-Learning, which is the modification of estimating Q-value by tabular or function approximation, is proposed to demonstrate the efficiency of reinforcement learning in simulations of a humanoid robot pushing a box. The box moves in the direction in which the robot asserts force. To push the box to the target point, the robot needs to learn how to adjust angles, avoid obstacles, and keep balance. Simulation results show the proposed method outperforms Q-Learning without using adaptive states.

This paper investigates the problem of robust stabilization for uncertain Takagi–Sugeno (T–S) fuzzy systems with additive time varying delays. An appropriate Lyapunov-Krasovskii function is considered for solving this problem, and obtains considerably less conservative results than existing methods. The proposed approach constructs a new Lyapunov-Krasovskii functional using two additive delay components, and no free weighting matrices are employed in the theoretical result derivation. This reduces the number of scalar decision variables in linear matrix inequalities. The fuzzy state feedback gain is derived through the numerical solution of a set of linear matrix inequalities (LMIs). Finally, numerical examples are provided to illustrate the effectiveness of the proposed method, and to allow comparison with previous works.

Background: The quality of construction is crucial in evaluating steel structure. However, traditional quality control methods for steel structure construction have been criticized for their lack of intelligence, resulting in a heavier reliance on manual experience and post-construction inspections to address quality issues. This shortcoming makes quality management inefficient and labor-intensive. To address this issue, this paper proposes a smart quality control method based on digital twin technology. Methods: In this framework, data collection is used for subsequent quality control throughout the construction process. To improve pre-construction quality control, a mixed reality (MR) system is used to guide and train personnel. During the steel structure construction process, the Markov method is used to analyze and predict real-time data. Results: To test the effectiveness of the proposed method, ten sets of parallel tests were conducted to predict whether the bolt torque value was normal or not, resulting in an 80% accuracy rate. Conclusions: The proposed method for steel structure construction quality control was effectively certified, achieving active prevention and real-time control of quality problems and improving the overall intelligence level of quality control.

Background: In recent years, with the development of the Internet of Vehicles, a variety of novel in-vehicle application devices have surfaced, exhibiting increasingly stringent requirements for time delay. Vehicular edge networks (VEN) can fully use network edge devices, such as roadside units (RSUs), for collaborative processing, which can effectively reduce latency. Objective: Most extant studies, including patents, assume that RSU has sufficient computing resources to provide unlimited services. But in fact, its computing resources will be limited with the increase in processing tasks, which will restrict the delay-sensitive vehicular applications. To solve this problem, a vehicle-to-vehicle computing task offloading method based on deep reinforcement learning is proposed in this paper, which fully considers the remaining available computational resources of neighboring vehicles to minimize the total task processing latency and enhance the offloading success rate. Methods: In the multi-service vehicle scenario, the analytic hierarchy process (AHP) was first used to prioritize the computing tasks of user vehicles. Next, an improved sequence-to-sequence (Seq2Seq) computing task scheduling model combined with an attention mechanism was designed, and the model was trained by an actor-critic (AC) reinforcement learning algorithm with the optimization goal of reducing the processing delay of computing tasks and improving the success rate of offloading. A task offloading strategy optimization model based on AHP-AC was obtained on this basis. Results: The average latency and execution success rate are used as performance metrics to compare the proposed method with three other task offloading methods: only-local processing, greedy strategy- based algorithm, and random algorithm. In addition, experimental validation in terms of CPU frequency and the number of SVs is carried out to demonstrate the excellent generalization ability of the proposed method. Conclusion: The simulation results reveal that the proposed method outperforms other methods in reducing the processing delay of tasks and improving the success rate of task offloading, which solves the problem of limited execution of delay-sensitive tasks caused by insufficient computational resources.