Transfer Components Research Articles

БАЛАНС И КИНЕМАТИКА МАССЫ ЛЕДНИКА ЦЕНТРАЛЬНЫЙ ТУЙЫКСУ, СЕВЕРНЫЙ СКЛОН ИЛЕ АЛАТАУ

Based on direct observations of the Central Tuiyksu glacier for the period 2006…2023, calculations of the components of external and internal mass transfer are presented and comparison of the obtained values for the periods 1958…1972 and 1977…1992 is provided. To determine the kinematic components of the mass balance, stakes drilled into the ice with geodetic reference on the ground were used. During the considered period, the glacier retreated by 420 meters, the mass balance on average was -0,54 m w.e., for the entire observation period, since 1958, the glacier retreated by 1190 m with the average mass balance value of -0,42 m w.e. for 1956…2023. The annual mass balance in the kinematic accumulation area for the period 2006...2023 was +0,13 m, for the same period the ice replacement from accumulation area to kinematic ablation area was -1,24 m, while kinematic accumulation area increased by 17 %, and kinematic ablation area decreased by 46,3 %. The inflow of ice from kinematic accumulation area compensated losses in kinematic ablation area in the period 1977/78…1991/92 only by half (49,5 %) and in the period from 2006/07 to 2022/23 already by 75,9 %. The maximum ice movement rates of the Central Tuiyksu glacier of 22…24 m/year were observed in the period 1956…65, they decreased to 16 m/year in the period 1977…1992 and in the period 2006…2023 were 12,8 m/year

DGA-Based Fault Diagnosis Using Self-Organizing Neural Networks with Incremental Learning

Power transformers are vital components of electrical power systems, ensuring reliable and efficient energy transfer between high-voltage transmission and low-voltage distribution networks. However, they are prone to various faults, such as insulation breakdowns, winding deformations, partial discharges, and short circuits, which can disrupt electrical service, incur significant economic losses, and pose safety risks. Traditional fault diagnosis methods, including visual inspection, dissolved gas analysis (DGA), and thermal imaging, face challenges such as subjectivity, intermittent data collection, and reliance on expert interpretation. To address these limitations, this paper proposes a novel distributed approach for multi-fault diagnosis of power transformers based on a self-organizing neural network combined with data augmentation and incremental learning techniques. The proposed framework addresses critical challenges, including data quality issues, computational complexity, and the need for real-time adaptability. Data cleaning and preprocessing techniques improve the reliability of input data, while data augmentation generates synthetic samples to mitigate data imbalance and enhance the recognition of rare fault patterns. A two-stage classification model integrates unsupervised and supervised learning, with k-means clustering applied in the first stage for initial fault categorization, followed by a self-organizing neural network in the second stage for refined fault diagnosis. The self-organizing neural network dynamically suppresses inactive nodes and optimizes its training parameter set, reducing computational complexity without sacrificing accuracy. Additionally, incremental learning enables the model to continuously adapt to new fault scenarios without modifying its architecture, ensuring real-time performance and adaptability across diverse operational conditions. Experimental validation demonstrates the effectiveness of the proposed method in achieving accurate, efficient, and adaptive fault diagnosis for power transformers, outperforming traditional and conventional machine learning approaches. This work provides a robust framework for integrating advanced machine learning techniques into power system monitoring, paving the way for automated, real-time, and reliable transformer fault diagnosis systems.

π-π stacking interactions in tryptophan-lumiflavin-tyrosine: a structural model for riboflavin insertion into riboflavin-binding protein.

Riboflavin (RF), also known as B2 vitamin, is the precursor to flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD), two co-enzymes involved in many electron transport processes. Interactions of the isoalloxazine ring, common to all three compounds, are of great interest due to their biological function in flavoproteins and relevance in the transport by the carrier protein leading to development of drug delivery strategies and non-invasive diagnostics techniques. Based on protein crystallographic data, a computational investigation of the interactions in the complexes between lumiflavin, a model compound, and aromatic amino acids, tyrosine and tryptophan, was pursued with the goal of characterizing noncovalent interactions. Density functional theory (DFT) served as the computation framework for all calculations, utilizing long-range corrected hybrid functionals LC-ωPBE and ωB97XD in conjunction with the 6-311+ +g** basis set. The solvation effects were incorporated through the implementation of the polarizable continuum model (PCM) simulating an aqueous solvent environment. The geometries of the five most stable complexes show exclusively p-p interactions among the aromatic moieties in a displaced parallel plane stacking arrangement with interplanar heights and displacements in the range of 3.22-3.62Å and 0.50-0.63Å, respectively, at ωB97XD level. The calculated total energies and binding energies indicate two stabilizing p-p interactions: lumiflavin-tyrosine and lumiflavin-tryptophan, with the later stronger for the more stable complexes by 2kcalmol-1. The complexes are less entropically favored than the independent molecules as verified by the positive association free Gibbs energies with LC-ωPBE and nearly zero with ωB97XD. Orbital analysis indicates a smaller HOMO-LUMO gap for complexes compared to the individual compounds suggesting a charge transfer component to the interaction. Moreover, the HOMO is localized on tryptophan and HOMO-1 on tyrosine, consistent with the strength of the respective interactions with lumiflavin. The initial geometry was based on the atom coordinates of the bonding tryptophan-riboflavin-tyrosine region in the protein crystallographic data with the ribityl tail being discarded, leading to a model complex: tryptophan-lumiflavin-tyrosine. The initial conformational search using the Amber force field within the Gabedit led to 30 unique conformations. The subsequent calculations, energy optimization and orbital analysis, were performed in Guassian16 at density functional theory (DFT) level, utilizing long-range corrected hybrid functionals LC-ωPBE and ωB97XD in conjunction with the 6-311+ +g** basis set. The solvent, water, was accounted for using the polarized continuum model (PCM).

Engineering mechanisms of proton-coupled electron transfer to a titanium-substituted polyoxovanadate-alkoxide.

Metal oxides are promising catalysts for small molecule hydrogen chemistries, mediated by interfacial proton-coupled electron transfer (PCET) processes. Engineering the mechanism of PCET has been shown to control the selectivity of reduced products, providing an additional route for improving reductive catalysis with metal oxides. In this work, we present kinetic resolution of the rate determining proton-transfer step of PCET to a titanium-doped POV, TiV5O6(OCH3)13 with 9,10-dihydrophenazine by monitoring the loss of the cationic radical intermediate using stopped-flow analysis. For this reductant, a 5-fold enhanced rate (k PT = 1.2 × 104 M-1 s-1) is accredited to a halved activation barrier in comparison to the homometallic analogue, [V6O7(OCH3)12]1-. By switching to hydrazobenzene as a reductant, a substrate where the electron transfer component of the PCET is thermodynamically unfavorable (ΔG ET = +11 kcal mol-1), the mechanism is found to be altered to a concerted PCET mechanism. Despite the similar mechanisms and driving forces for TiV5O6(OCH3)13 and [V6O7(OCH3)12]1-, the rate of PCET is accellerated by 3-orders of magnitude (k PCET = 0.3 M-1 s-1) by the presence of the Ti(iv) ion. Possible origins of the accelleration are considered, including the possibility of strong electronic coupling interactions between TiV5O6(OCH3)13 with hydrazobenzene. Overall, these results offer insight into the governing factors that control the mechanism of PCET in metal oxide systems.

Transfer Learning Estimation and Transferability of LNC and LMA Across Different Datasets

Leaf mass per area (LMA) and leaf nitrogen concentration (LNC) are both essential parameters in plant ecology, which can reflect the growth status of plants. The features of LMA and LNC can be captured using spectral reflectance in a remote sensing approach. While the relationships between spectra and leaf trait variance across different species with estimation performance are unclear, the development of assessment and transferable models to predicate LMA and LNC are prevented. Hence, we analyzed the variance of raw spectra and spectral data difference with four pretreated approaches (SG—Savitzky–Golay filter, SNV—Standard Normalized Variate, MSC—Multiplicative Scatter Correction analysis, and normalize), LMA, and LNC over six remote sensing datasets by a transfer component analysis (TCA) approach. Spectra combined with the Successive Projections Algorithm (SPA) were also presented to extract wavelengths with higher important coefficients to minimize the redundancy of datasets. The variance of normalized spectra between different datasets showed a minor degree of variance, and LNC spectra variance was decreased by the SPA. The results also showed that a smaller LMA and LNC variance is presented over different datasets when the trait values with higher distribution probabilities are close to each other. The LNC and LMA estimation performance in transfer models established by partial least squares regression (PLS), support vector regression (SVR), extreme gradient boosting (XGB), and random forest regression (RFR) algorithms across different datasets were employed, in which the RFR transfer models performed good prediction results. The relationships between spectra and leaf trait variance and estimation performance in RFR transfer models over different datasets were evaluated. LMA distance has a significant influence on estimation performance in the transfer model, and the variance of spectra with all pretreated approaches showed a very significant effect on LNC accession performance. Furthermore, we proposed a weight coefficient of spectral data updating combined with the TCA and RFR approach (WDT-RFR) transfer model to improve transferability between datasets and promote estimation performance in the transfer model. Compared to the RFR transfer model using spectra without updating, the root mean square error (RMSE) of the WDT-RFR transfer model with 5% samples transferred to estimate LMA and LNC increased by 7.9% and 4.8% on average, respectively. The estimation results showed that our transfer model showed a superior estimation performance.

Neutrinos of magnetorotational supernovae

The neutrino dynamics in hot and dense magnetized matter corresponding to a supernova explosion is considered. It is shown that accounting for fluctuations during interaction of neutrinos with matter leads to the Fokker–Planck equation for the dynamics of the phase space distribution function. The addition to the energy transfer component of the kinetic equation is determined by straggling in neutrino collisions with a magnetized nucleon gas caused by the neutral current Gamow–Teller interaction. When accounting for the effect of fluctuations, the switching of acceleration and stopping regimes in neutrino evolution is evident for average energy. The effect of fluctuations leads to an additional increase in the hardness of the neutrino spectra. It is shown that the high-energy component of the electron antineutrino flux is enhanced in addition due to the effect of neutrino oscillations. Such a strengthening of the spectrum hardness is especially noticeable in the case of the inverted mass ordering and makes the signal more registrable by ground-based detectors. The possibilities of detecting supernova neutrinos by KM3NeT and Baikal-GVD observatories are discussed. The sensitivity of counting rate to the mass ordering is demonstrated to increase at growing difference in the hardness of energy spectra for various flavors.

Characterisation of the Cytochrome P450 Monooxygenase CYP116B46 from Tepidiphilus thermophilus as a Homogentisic Acid Generating Enzyme and its Conversion to a Peroxygenase.

The heme enzymes of the cytochrome P450 superfamily (CYPs) catalyse the selective hydroxylation of unactivated C-H bonds in organic molecules. There is great interest in applying these enzymes as biocatalysts with a focus on self-sufficient CYP 'fusion' enzymes, comprising a single polypeptide chain with the electron transfer components joined to the heme domain. Here we elucidate the function of the self-sufficient CYP116B46 fusion enzyme, from the thermophilic bacterium Tepidiphilus thermophilus. We demonstrate that it efficiently hydroxylates aromatic organic acids, exemplified by oxidation of 2-hydroxyphenylacetic acid to homogentisic acid (2,5-dihydroxyphenylacetic acid), an important metabolite in bacterial catabolism. In line with the thermophilic nature of the source bacterium, activity increased at higher temperatures, (50 °C), with a catalytic preference for NADPH over NADH. While self-sufficient fusion enzymes simplify biocatalysis; engineered peroxygenase activity is also a key advance in the application of these enzymes as biocatalysts as it eliminates the need for electron transfer partner proteins and nicotinamide cofactors. We demonstrate that a T278E mutation in the heme domain of CYP116B46, confers peroxygenase activity. This engineered peroxygenase enzyme is stable to elevated temperatures and catalytic concentrations of hydrogen peroxide, with an observed optimal activity resulting in a total turnover number of ~650.

Photorealistic attention style transfer network for architectural photography photos

Architectural photography style transfer, a task in computer vision, employs deep learning algorithms to transform the style of architectural photograph while preserving key structure and content. Existing algorithms face challenges due to the intricate details of buildings, including diverse shapes, lines, and textures. Moreover, considerations for artistic effects in architectural photography style transfer, such as lighting, shadows, and atmosphere, require high-quality image generation algorithms. However, current algorithms often struggle to address these complexities, leading to loss or blurring of details and less realistic images. To overcome these challenges, this paper proposes a Photorealistic Attention Style Transfer Network. The proposed approach utilizes a semantic segmentation model to accurately segment the input image into foreground and background components for independent style transfer. Subsequently, the transferred images are refined by focusing on intricate building parts using the coordinate attention mechanism. Additionally, the network incorporates loss functions to capture light, shadow, and colors in stylish images, ensuring realism while maintaining aesthetic appeal. Through comparative experiments, the proposed network shows better performance in terms of image fidelity and overall aesthetics, and the SSIM and PSNR indices are also better than the current mainstream methods.

Theoretical and FEA study on shear performance of novel shear connectors

Shear connectors are essential components for force transfer in composite beams. Compared to conventional welded-bolt connectors, bolted shear connectors offer the advantage of easy removal. However, traditional bolted connectors for steel-concrete composite beams feature threads on the contact surface, reducing the shear cross-sectional area of the bolts. To address this limitation, this paper proposes a novel type of friction-clamping bolt connector, referred to as Connecting Lock Combination Bolts (CLCB11CLCB:Connecting Lock Combination Bolts；). The finite element software ABAQUS was used to develop an analytical model for the CLCB, and a comparative analysis was conducted with traditional connectors. The static performance of the new shear connectors was evaluated, and a shear load capacity formula applicable to CLCB connectors was derived. Additionally, linear and exponential functions were employed to fit the load-slip curve of the CLCB connectors. The results indicate that the shear load capacity of CLCB connectors exceeds that of conventional bolted connectors. Moreover, the bolt diameter and concrete strength are the primary factors influencing the ultimate shear load capacity and stiffness of CLCB connectors. The preload on the bolts enhances the shear stiffness, while parameters such as bolt hole clearance, the height of the lower nutted rod, and the clearance between the upper and lower bolt rods are crucial in ensuring the proper functioning of CLCB connectors. Furthermore, the shear capacity formula provides an accurate prediction of CLCB connector performance with an error margin of less than 6 %, and the load-slip curve model effectively illustrates the load-slip relationship under shear.

Enhanced prediction of river dissolved oxygen through feature- and model-based transfer learning

Water quality monitoring data from various points within the same basin often show non-uniformity. A key scientific question is how to extract relevant knowledge from data-rich sites (source domains) and leverage the possible inter-site consistency of water quality to compensate for the limitations of data-poor sites (target domains). Transfer learning (TL) methods can improve the applicability of water quality predictions for data-poor sites but their comparison and combination have not been fully explored. This study employs feature-based (Transfer Component Analysis, TCA) and model-based (pretraining and fine-tuning) transfer learning, to assist in constructing Long Short-Term Memory (LSTM) models for forecasting the dissolved oxygen (DO) levels in the West Channel of Guangzhou, southern coastal China. The LSTM models at Yagang and Shimen stations were constructed as the basic and baseline models for source and target domains, respectively. By comparing and selecting different transfer learning strategies, the best single-type TL strategy emerged as a multi-sequence LSTM model without TCA but with the fully connected layer frozen after pretraining. It achieved increases in validation Nash efficiency coefficient (NSE) of 5.2%, 10.8%, and 46.2% for predicting DO over the next 3 days, respectively, compared to the baseline LSTM model at Shimen station. The best combined TL strategy involved using TCA and freezing the second fully connected layer in a multi-sequence LSTM model. It improved upon the baseline LSTM model with a validation NSE increase of 5.3%, 21.4%, and 48.7% over the next three days, respectively. This study demonstrates that combining feature- and model-based transfer learning methods can yield better DO prediction performance in data-poor rivers than using a single-type transfer learning method.

Experimental Investigation of Heat Pipe Flow Dynamics and Performance

Due to their safety, efficiency, and passive operation, heat pipes have found diverse applications that include nuclear microreactors. Heat pipes enable increased reliability in microreactors, as they eliminate the need for reactor coolant pumps and their associated auxiliary systems while resulting in a greatly reduced spatial footprint. Experimental work is needed to support and expedite the design and licensing of heat pipe microreactors, especially the validation of heat pipe performance, as key heat transfer components. The present work develops a comprehensive heat pipe experimental database covering a wide range of heat pipe operating conditions. In addition, two-phase thermosyphon experiments are conducted to serve as a benchmark for performance. The operating conditions are determined based on previously developed scaling laws for heat pipes and two-phase thermosyphons using low-temperature working fluids. The tested heat pipe is about 2 m long and equipped with in-house-developed annulus screen wicks. To allow for the investigation of heat pipe flow dynamics, various instruments are incorporated to acquire heat pipe pressures, pressure drops, and temperatures. In particular, a fiberoptic sensor is implemented to measure temperatures along the centerline of the entire heat pipe. The results can be directly applied to the advancement of numerical tools currently under development for heat pipe microreactor analysis.

Impact of Conditional Cash Transfer Program on Household Welfare in Southern Ethiopia: A matching estimator approach

Cash transfers (CTs) are now high on the agenda of most governments in low- and middle-income countries. Chronic food insecurity is one of the problems that has plagued millions of Ethiopians for centuries. This study evaluated the impact of participation in the conditional cash transfer (CCT) component of the PSNP (Public Works or PW) on household welfare in Karat Zuria District, Konso Zone, Southern Ethiopia. The cross-sectional survey data was gathered from 244 sample households, which were taken from beneficiary and control groups by using a multistage sampling procedure. Propensity score matching (PSM) was used to evaluate the program's impact on livestock holdings and total income. The result obtained from logit regression indicated that participation in a conditional cash transfer program was negatively influenced by access to credit and cultivated land size, whereas it was positively influenced by shock experience and an iron-roofed house. The econometrics model result shows that, as compared to control households, the program intervention increased the beneficiary households’ livestock holding (measured in TLU) by 24.8 percent (0.53 TLU) and total income by 33 percent (1705.095 ETB), respectively. This study demonstrated how the program had a considerable impact on household livestock holdings and total income.

Thermal processes affected by carbon dioxide near ground surface

This study provides penetrative investigation of the effects of carbon dioxide concentration on time-dependent temperature and energy fluxes on the ground surface subject to solar irradiation. While global warming significantly impacts human life, the factors responsible remain controversial. This work considers unsteady one-dimensional heat conduction and radiative heat transfer, including collimated and diffuse components as functions of longitude, latitude, and altitude. Diffuse radiation depends on the different absorption bands of carbon dioxide and water vapor, as functions of wavelength, temperature, concentrations or pressure. The predicted results using COMSOL computer code show that the effects of carbon dioxide concentration on ground surface temperature are negligibly small, for example, over a 5-year period. Time-dependent ground surface temperature strongly depends on the absorption or dissipation of diffuse radiation and heat conduction. Diffuse radiation has a damping effect on temperature variation. Temperature changes due to solar irradiation absorption in the atmosphere are also negligibly small, despite solar irradiation being much greater than diffuse radiation and heat conduction. The absorption or dissipation of diffuse radiation depends on the dominant absorption bands centered at 4.3 and 15 μm of carbon dioxide at different times. This study, from the viewpoint of energy conservation, identifies diffuse radiation as a critical factor influencing the rate of temperature change at the ground surface, without assuming radiative or thermal equilibrium or modeling convection. Poor management of this radiation would hinder efforts to avoid droughts, water scarcity, severe fires, rising sea levels, flooding, catastrophic storms, and biodiversity loss.

Feature Transfer Learning for Fatigue Life Prediction of Additive Manufactured Metals With Small Samples

ABSTRACTA feature transfer learning (FTL)‐based model is proposed to address small‐sample problems in fatigue life prediction of additively manufactured (AM) metals. Transfer component analysis (TCA) is studied for data alignment before model training. Correspondingly, two TCA improvement strategies are further considered to aggregate training data from distinct AM processing conditions. An experimental database consisting of 103 fatigue data is built for model evaluation. The results demonstrate that the proposed model outperforms conventional machine learning models and other transfer learning‐based models in terms of accuracy and data demand, showing good applicability for AM fatigue life assessment.

Similarity analysis to enhance Transfer Learning for Damage Detection

Abstract One significant challenge in machine learning for Structural Health Monitoring (SHM) is reusing previously trained classifiers. A classifier might be suitable for one situation but not for another. Transfer learning techniques try to overcome this difficulty. In SHM, it is common to use the modal parameters as features; however, they are highly influenced by boundary conditions, geometry, and the level of structural damage. This work proposes an approach that performs a similarity analysis to select features before applying transfer learning, aiming at improving classification and damage detection. The reasoning is that a higher similarity leads to a more efficient transfer of learning and, consequently, a better classification. Transfer learning is conducted via the domain adaptation technique known as Transfer Component Analysis (TCA), and cases with low similarity are compared to those with high similarity. Two datasets are analyzed. The first consists of a beam under different boundary conditions, and data is generated through numerical simulations. The second derives from an experimental setup of bolted joints with loosening damage. The proposed strategy, which uses a cosine-type similarity, is shown to improve the transfer learning classification.

Implementing a chronic disease self-management program in Moldova: a roadmap for sustainability

Abstract Chronic disease self-management programs are evidence-based interventions but achieving their sustainability can be challenging. We propose a roadmap to sustainability that examines the transferability of intervention components and fosters multi-level and multisectoral partnerships. Since 2019, the Swiss Agency for Development and Cooperation’s Healthy Life project, in cooperation with the Ministry of Health of Moldova, has implemented a self-management program in rural communities. At pilot stage, we achieved a 53% increase in participants’ self-efficacy (Palmeirim et al, 2024). To date, the program has reached 1432 beneficiaries in 70 communities. The intervention cost per participant was approximately 74 Euros. Our evidence supported the inclusion of self-management in the national non-communicable disease program 2023-2027. We designed a roadmap to hand over this innovation to Moldovan public health institutions. Firstly, we unpacked the intervention into transferable resources, processes and roles. These included human resources (a network of 140 facilitators), pedagogical resources (culturally adapted facilitator training and self-management workshop content) and key operational processes (coordination, resource management, partnership building, monitoring and evaluation). Then, we convened a national workshop to discuss the transfer of these components to relevant institutions. We emphasized working in a coordinated partnership across institutions for intervention delivery. Finally, we launched a national working group that will work on the operational modalities of this multi-stakeholder partnership. Achieving an enabling policy environment for self-management created favorable conditions and accountability for institutional partnership building. A strategic reflection on our partner ecosystem was key to leverage local and national interests in continuing self-management approaches. Key messages • We have established priorities among transferable intervention components to match country stakeholders’ ability and readiness to sustain innovative approaches. • We are building an institutional partnership across sectors based on complementarity of experience to sustain the delivery of a self-management intervention.

A transfer learning method for near infrared models of potato starch content and traceability from different origins

The robustness of near-infrared (NIR) models in detecting agricultural product quality is challenged by the differences in statistical conditions such as territory, variety, and collection time, and geographical differences in sample sources. This study aimed to use global and local migration models to improve the generalization ability of prediction models of potato starch content from different sources. The results showed that both models could eliminate the influence of samples from different sources on the model performance; the transfer component analysis (TCA)-based model was superior to the global model. The correlation coefficient (RP), root-mean-square error of prediction (RMSEP), and relative percent deviation (RPD) of the prediction model of starch content in the target domain of Whale Optimization Algorithm (WOA)–Radial Basis Function (RBF) based on the TCA method reached 0.931, 0.763 %, and 2.740, respectively. After the second model transfer, the model still had an extremely reliable performance (RPD = 2.050 > 2). The precision and accuracy of the WOA–RBF traceability model reached 91.25 % and 95 %, respectively. This study provided a feasible solution to the problem of poor generalization ability of a single-source model and proposed an effective, stable, and universal method for nondestructive testing of potato traceability.

Dynamic Temporal Denoise Neural Network with Multi-Head Attention for Fault Diagnosis Under Noise Background.

Fault diagnosis plays a crucial role in maintaining the operational safety of mechanical systems. As intelligent data-driven approaches evolve, deep learning (DL) has emerged as a pivotal technique in fault diagnosis research. However, the collected vibrational signals from mechanical systems are usually corrupted by unrelated noises due to complicated transfer path modulations and component coupling. To solve the above problems, this paper proposed the dynamic temporal denoise neural network with multi-head attention (DTDNet). Firstly, this model transforms one-dimensional signals into two-dimensional tensors based on the periodic self-similarity of signals, employing multi-scale two-dimensional convolution kernels to extract signal features both within and across periods. Secondly, for the problem of lacking denoising structure in traditional convolutional neural networks, a temporal variable denoise (TVD) module with dynamic nonlinear processing is proposed to filter the noises. Lastly, a multi-head attention fusion (MAF) module is used to weight the denoted features of signals with different periods. Evaluation on two datasets, Case Western Reserve University bearing dataset (single sensor) and Real aircraft sensor dataset (multiple sensors), demonstrates that the DTDNet can reduce the useless noises in signals and achieve a remarkable improvement in classification performance compared with the state-of-the-art method. DTDNet provides a high-performance solution for potential noise that may occur in actual fault diagnosis tasks, which has important application value.

Unsupervised domain adaptation for drive-by condition monitoring of multiple railway tracks

Monitoring railway tracks through drive-by vibration data collected by in-service trains offers a cost-effective and adaptable solution for inspecting multiple railway lines. However, numerous existing drive-by monitoring methods rely on supervised learning models, necessitating extensive labelled data for each line. In this paper, a novel framework is proposed based on Unsupervised Domain Adaptation (UDA) concept which facilitates the transfer of a geometric defects diagnosis model learned from one line to a new line without the need for any labelled data from the new line. The proposed framework learns the dynamic-based features that are sensitive to damage and also invariant to different railway tracks. It comprises three components: data pre-processing, UDA implementation, and damage diagnosis. The framework uses the data from the source domain, including corresponding labels, as well as the unlabelled data from the target domain as input. The outputs of the framework consist of the predicted labels for the target domain. The performance of the proposed framework is evaluated using a comprehensive dataset of field measurements of a high-speed train passing 4 different lines within the French high-speed rail network. The proposed UDA framework is implemented using four common UDA algorithms including Information-Theoretical Learning (ITL), Geodesic Flow Kernel (GFK), Transfer Component Analysis (TCA), and Subspace Alignment (SA). The results show that the proposed framework has a 14% increase in the anomaly detection accuracy compared to traditional unsupervised learning methods in which UDA is not used. Furthermore, this study investigates the impact of incorporating a percentage of target data labels during training (semi-supervised domain adaptation), along with various sensor layouts and different tuning parameters, on the accuracy of the proposed approach. The results show that the proposed framework can significantly facilitate the monitoring of railway track conditions using the data collected by in-service trains which could be great interest of railway owners.

Individually Tunable Energy Levels of Oligomers Based on N-B←N Units.

Opto-electronic properties and device performance of organic semiconductors are mainly determined by energy levels of their frontier molecular orbitals, e.g. lowest unoccupied molecular orbital (ELUMO) and highest occupied molecular orbital (EHOMO) in the ground state, first singlet state (ES1) and first triplet state (ET1) in the excited state. These energy levels are always intricately intertwined. Herein, we report a series of monodisperse oligomers based on double B←N bridged bipyridine (BNBP) units. With the increasing number of repeating units, the oligomers exhibit gradually downshifted ELUMO and nearly unchanged EHOMO due to the different distribution of the frontier molecular orbitals of the oligomers. Moreover, the oligomers exhibit gradually decreasing ES1 and nearly unchanged ET1 because of the different contributions of the charge transfer component in the excited state. This work provides new insight into energy level tuning of organic semiconductors, which is important for high-performance organic opto-electronic devices.