Abstract ：

The poor adhesion of asphalt-granite limits the application of granite. This research aims to enhance the adhesion of asphalt and granite by establishing chemical connections between the hydroxyl groups on granite surface and the incompletely reacted isocyanate groups in polyurethane (PU) modified asphalt. For this purpose, end-capped PU modified asphalt (PUMA) was prepared, the adhesion mechanism between PUMA and granite was analyzed by Raman spectroscopy, the adhesion properties between PUMA and granite at different curing periods were evaluated by boiling and pull-out tests, and the wettability of PUMA on granite and its moisture sensitivity were evaluated using surface free energy (SFE) method. The findings suggested that the incompletely reacted isocyanate groups in PUMA can establish a chemical connection with the hydroxyl groups on the surface of granite. The adhesion between asphalt and granite can be greatly improved by the addition of PU, and this improvement is amplified as the curing time is extended. When the curing time is extended to 3 days, the fracture mode between asphalt and granite changed from adhesive failure to cohesive failure, and after 15 days of curing, the coating ratio of PUMA on granite following the water boiling test can be raised by 28.11% in comparison to the base asphalt. The water damage resistance of PUMA-granite system can be increased by adding PU and the adhesion work and combination ratio of the freshly prepared PUMA were increased by 55.91% and 79.45%, respectively, as compared to base asphalt. This study has great significance for enhancing the adhesion between asphalt and granite and encouraging granite application.

Keyword ：

Granite Polyurethane modified asphalt (PUMA) Interfacial adhesion End-capped polyurethane

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Objective. Electroencephalogram (EEG) analysis has always been an important tool in neural engineering, and the recognition and classification of human emotions are one of the important tasks in neural engineering. EEG data, obtained from electrodes placed on the scalp, represent a valuable resource of information for brain activity analysis and emotion recognition. Feature extraction methods have shown promising results, but recent trends have shifted toward end-to-end methods based on deep learning. However, these approaches often overlook channel representations, and their complex structures pose certain challenges to model fitting. Approach. To address these challenges, this paper proposes a hybrid approach named FetchEEG that combines feature extraction and temporal-channel joint attention. Leveraging the advantages of both traditional feature extraction and deep learning, the FetchEEG adopts a multi-head self-attention mechanism to extract representations between different time moments and channels simultaneously. The joint representations are then concatenated and classified using fully-connected layers for emotion recognition. The performance of the FetchEEG is verified by comparison experiments on a self-developed dataset and two public datasets. Main results. In both subject-dependent and subject-independent experiments, the FetchEEG demonstrates better performance and stronger generalization ability than the state-of-the-art methods on all datasets. Moreover, the performance of the FetchEEG is analyzed for different sliding window sizes and overlap rates in the feature extraction module. The sensitivity of emotion recognition is investigated for three- and five-frequency-band scenarios. Significance. FetchEEG is a novel hybrid method based on EEG for emotion classification, which combines EEG feature extraction with Transformer neural networks. It has achieved state-of-the-art performance on both self-developed datasets and multiple public datasets, with significantly higher training efficiency compared to end-to-end methods, demonstrating its effectiveness and feasibility.

Keyword ：

self-attention mechanism. electroencephalographic deep learning power spectral density emotion recognition

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Load and resistance factor design (LRFD) is a suitable format for the reliability-based limit state design of structures. It has been adopted in many countries, such as the United States, Europe, Canada, and Japan. Usually, the first-order reliability method (FORM) is used to estimate the load and resistance factors, but it requires the determination of design points and complicated double iterative computations. Therefore, FORM is not easy or practical for engineers to use. This paper presents a simple, accurate method to determine the load and resistance factors utilizing the three-moment transformation, which does not require derivative-based iterations and can estimate the load and resistance factors without using the distribution of random variables. In addition, the proposed method provides enough accurate results within a wide range of target reliability indices. Therefore, this method should be effective and convenient for determining the load and resistance factors in actual practice. Three numerical examples illustrate the proposed method's efficiency and accuracy; FORM provides a benchmark for comparison. © 2022 ISRERM Organizers. Published by Research Publishing, Singapore.

Keyword ：

Structural design

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Optimizing the flow channel structures can improve the efficiency of proton exchange membrane fuel cells and promote the efficient utilization of clean energy. This study investigates the impacts of rectangular-baffle and tapered flow channels on electrical performance, temperature distribution and mass transfer of proton exchange membrane fuel cells, based on a 3D, multi-physical model. The equivalent average depth is adopted as the reference benchmark to mitigate the influence of channel depth. Results indicate that adopting variable-depth channels enhances convective mass transfer on the porous layer surface under higher voltage. The convective mass transfer of oxygen is dominant in the activation polarization region, while diffusion is dominant in the concentration polarization region. Variable-depth channels enhance heat and mass transfer in comparison to the straight channel. The tapered channel can improve the mass transfer flux in oxygen starvation region, and the drainage ability of the rectangular baffle channel is inferior to the tapered channel. The rectangular baffle channel demonstrates better convective heat transfer performance and temperature uniformity. A rectangular-baffle channel can increase net power by 4.31%, while a tapered channel can achieve a maximum increase of 9.27% when the pumping power is considered. Therefore, incorporating a tapered channel is an optimal strategy to boost cell efficiency.

Keyword ：

net power proton exchange membrane fuel cell tapered flow channel mass transfer Rectangular baffle

Cite：

Copy from the list or Export to your reference management。

Abstract ：

As business rapidly evolves, the types of business services become increasingly complex. The conventional approach of manually configuring rules for maintenance and quality assurance systems is gradually proving inadequate, and there is a growing demand for intelligent operations (AIOps) and quality assurance. This paper proposes an anomaly detection algorithm by ensemble various algorithms. Originating from operation and maintenance scenarios, this algorithm is gradually applied to business scenarios. The process involves streaming data collection, data preprocessing, anomaly detection models, and monitoring alerts, thereby realizing multidimensional anomaly detection in both operational and business scenarios. In comparison to traditional monitoring platforms, the proposed algorithm saves manpower and exhibits higher accuracy. Furthermore, when compared to traditional machine learning algorithms, it saves resources and provides faster response times. © 2024 IEEE.

Keyword ：

Learning algorithms Anomaly detection Machine learning Quality assurance Data acquisition

Cite：

Copy from the list or Export to your reference management。

Abstract ：

This study focuses on improving financial efficiency and rationally allocating financial resources in the Beijing-Tianjin-Hebei, based on data from 2011 to 2019. Firstly, the Slack-Based Measure (SBM) model based on Shannon Entropy is adopted to measure financial efficiency in Beijing-Tianjin-Hebei. Secondly, a Beijing-Tianjin-Hebei inter-regional financial resource allocation scheme is proposed using the Generalized Equilibrium Efficient Frontier Data Envelopment Analysis (GEEFDEA) model. Finally, the external environmental factors influencing financial efficiency are explored using spatial and other econometric models. Efficiency measurement reveals that financial efficiency in Beijing-Tianjin-Hebei is unevenly developed. The financial efficiency of Beijing, Tianjin, and Zhangjiakou is high and stable, whereas the financial efficiency of Tangshan, Langfang, and Shijiazhuang is high, but fluctuates significantly. In comparison, the financial efficiency of other cities is low. The factors and the adjustment amount to improve urban financial efficiency are obtained by studying the allocation of financial resources. And it is found that Tianjin and Tangshan have more financial resource surpluses than other cities. By exploring influencing factors, it is found that the financial development level, innovation level, and infrastructure construction level of Beijing-Tianjin-Hebei significantly affect financial efficiency.

Keyword ：

GEEFDEA model SBM model Financial efficiency financial resource allocation

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Accurately identifying the unknown parameters of the Gurson-Tvergaard-Needleman (GTN) damage model is essential to better understand the damage behavior of materials and improve the simulation accuracy. In order to solve the problem that traditional parameter identification methods are cumbersome and inefficient, this paper proposed an efficient damage parameter identification strategy with the combination of microscopic analysis, finite element simulation, and response surface analysis. The relationship between the evolution of microvoids and plastic strain in DP800 steel was quantified through the in-depth analysis of the microstructure of tensile specimens. Response surface method and Box-Behnken experimental design were adopted to efficiently identify and optimize the crucial damage parameters (fN, fC, fF) with the difference ratio of fracture strain as the response variable. Through analysis of variance, the factors that have significant influence on the damage parameters were further confirmed. Through the experimental comparison with the simulation results based on the GTN damage model, the effectiveness of the identified damage parameters was verified, and the influence of these parameters on the stress-strain curve was deeply revealed. The results show that the damage parameter identification strategy has achieved remarkable results in accuracy and efficiency, and has important engineering significance for the accurate prediction of the actual dual-phase steel sheet forming process.

Keyword ：

Microstructure analysis Response surface methodology (RSM) Parameter identification GTN damage model Dual -phase steel

Cite：

Copy from the list or Export to your reference management。

Abstract ：

The objective of this study is to propose an interactive training method based on the Unity 3D platform and to evaluate its effectiveness in comparison to a video-based training method in curbing speeding behavior and promoting road safety. Based on the attributes unique to drivers with Type A behavior patterns, this study selected 36 drivers with Type A personality and assigned them to one of two groups (the Unity-based training group, UTG; and the video-based training group, VTG). Moreover, a test questionnaire consisting of 20 items was designed to reveal the drivers' mastery of knowledge related to speeding behavior through training. The research preliminarily analyzes the effectiveness of the interactive training method based on Unity 3D platform to improve speed management behavior by assessing the effects of the groups (the UTG and the VTG), test sessions (baseline, 30 min post-training, and three days post-training), and the interactive effects of the two factors on the test questionnaire score, the frequency of speeding and (FS), the percentage of time speeding (PTS), and the mean speed based on the two-way repeated-measures ANOVA. The results demonstrated that the Unity-based training method was, to some extent, more effective than the passive video training method at mastering knowledge related to speeding behavior and improving speed management behavior. Specifically, FS and PTS were significantly lower for the UTG than the VTG at the 30 min post-training sessions (p < 0.01) and the three-day post-training sessions (p < 0.01).

Keyword ：

Speeding behavior Unity 3D platform Traffic safety Driving education and training Interactive training method

Cite：

Copy from the list or Export to your reference management。

Abstract ：

To satisfy the differentiated service requirements of delay-sensitive and computing-intensive tasks in unmanned aerial vehicle (UAV) networks, it is urgent to efficiently allocate limited network resources to improve network performance. In this paper, we propose an intelligent task offloading scheme to optimize resource allocation in UAV networks with content caching. Specifically, we formulate the joint optimization of task offloading and resource allocation as a latency minimization model for the caching-assisted UAV system. Then, a new deep reinforcement learning (DRL) algorithm is designed to make offloading and resource allocation decisions based on current network state information, significantly improving resource utilization. Numerical results indicate that the model significantly reduces network latency in comparison to its existing benchmarks in caching-assisted UAV networks.

Keyword ：

content caching task offloading resource allocation unmanned aerial vehicle networks deep reinforcement learning

Cite：

Copy from the list or Export to your reference management。

Abstract ：

Connected vehicles (CVs) have demonstrated significant potential for addressing traffic issues. This paper presents a mathematical framework for stochastic heterogeneous traffic flow, integrating regular vehicles (RVs) and CVs, considering stochasticity, driver's dynamic time headway (DTH) characteristics, and the information flow topology (IFT) of CVs. We develop a novel car -following model (CFM) for RVs, accounting for both driver's stochasticity and DTH characteristics. Furthermore, we propose a dynamic model for CVs by integrating connected assisted driving strategies (CADS) into the RVs' model, which includes a dynamic information flow topology (DIFT) based on the time headway (TH) between vehicles within the communication range. We derive second -order exponential stability conditions for both RVs and CVs by employing the Lyapunov stochastic stability theory. We investigate the impact of driver stochasticity, DTH characteristics, and CADS on heterogeneous traffic flow characteristics through extensive numerical experiments. Model calibration results indicate that, in comparison to the state-of-the-art model, the proposed model exhibits superior prediction accuracy, achieving a 9.09% improvement at the group -driver level and a 10.47% improvement at the individual -driver level. Theoretical and numerical experimental results demonstrate that CVs with the proposed assisted driving strategy effectively mitigate traffic oscillations, and traffic flow stability improves as the CV penetration rate increases. Moreover, CVs can efficiently suppress stochasticity in traffic flow, with the strength of traffic fluctuations decreasing as the CV penetration rate grows. Different CV spatial distributions result in different propagation strengths of disturbances in traffic flow, adhering to a specific dual Gaussian distribution. Under various conditions, the average decline rates of speed fluctuations and energy consumption for traffic with the increasing CV penetration rate are 20%-50% and 10%-30%, respectively.

Keyword ：

Connected vehicles Stochastic linear stability Driver's stochasticity Dynamic time headway Heterogeneous traffic flow Dynamic information flow topology

Cite：

Copy from the list or Export to your reference management。