Search Results (171)

Accurate heat load forecasting is essential for the efficiency of District Heating Systems (DHS). Still, it is challenged by the need to model long-term temporal dependencies and nonlinear relationships with weather and other factors. This study proposes a hybrid deep learning framework combining an attention-enhanced Neural Basis Expansion Analysis for Time Series (N-BEATS) model and eXtreme Gradient Boosting (XGBoost). The N-BEATS component, with a multi-head self-attention mechanism, captures temporal dynamics, while XGBoost models non-linear impacts of external variables. Predictions are integrated using an optimized weighted averaging strategy. Evaluated on a dataset from 103 heating units, the model outperformed 13 baselines, achieving an MSE of 0.4131, MAE of 0.3732, RMSE of 0.6427, and R² of 0.9664. This corresponds to a reduction of 32.6% in MSE, 32.0% in MAE, and 17.9% in RMSE, and an improvement of 5.1% in R² over the best baseline. Ablation studies and statistical tests confirmed the effectiveness of the attention mechanism and ensemble strategy. This model provides an efficient solution for DHS load forecasting, facilitating optimized energy dispatch and enhancing system performance. Full article

The advent of quantum computing has introduced new opportunities for enhancing classical machine learning architectures. In this study, we propose a novel hybrid model, the HQDNN (Hybrid Quantum–Deep Neural Network), designed for the automatic detection of fake news. The model integrates classical fully connected neural layers with a parameterized quantum circuit, enabling the processing of textual data within both classical and quantum computational domains. To assess its effectiveness, we conducted experiments on the widely used LIAR dataset utilizing Term Frequency–Inverse Document Frequency (TF-IDF) features, as well as transformer-based DistilBERT embeddings. The experimental results demonstrate that the HQDNN achieves a superior recall performance—92.58% with TF-IDF and 94.40% with DistilBERT—surpassing traditional machine learning models such as Logistic Regression, Linear SVM, and Multilayer Perceptron. Additionally, we compare the HQDNN with SetFit, a recent CPU-efficient few-shot transformer model, and show that while SetFit achieves higher precision, the HQDNN significantly outperforms it in recall. Furthermore, an ablation experiment confirms the critical contribution of the quantum component, revealing a substantial drop in performance when the quantum layer is removed. These findings highlight the potential of hybrid quantum–classical models as effective and compact alternatives for high-sensitivity classification tasks, particularly in domains such as fake news detection. Full article

The aim was to investigate the impact of laser parameters on the surface morphology of ablated graphene and elucidate the interaction mechanism between carbon materials and femtosecond lasers. A pulsed laser with a wavelength of 1030 nm is employed to infer the ablation threshold of the surface and interface of graphene coatings formed through ultrasonic spraying. The ablation threshold of the coating–substrate interface is verified by numerical simulation. Incorporating the data of groove width and depth obtained from a three-dimensional profilometer and finite element simulation, an in-depth analysis of the threshold conditions of laser ablation in coating materials is accomplished. The results indicate that when the femtosecond laser frequency is 10 kHz, the pulse width is 290 fs, and the energy density reaches 0.057 J/cm², the graphene material can be effectively removed. When the energy density is elevated to 2.167 J/cm², a complete ablation of a graphite coating with a thickness of 1.5 μm can be achieved. The findings of this study validate the evolution law and linear relationship of ablation crater morphology, offering new references for microstructure design and the selection of controllable laser processing parameters. Full article

Deep learning has advanced medical image segmentation, yet existing methods struggle with complex anatomical structures. Mainstream models, such as CNN, Transformer, and hybrid architectures, face challenges including insufficient information representation and redundant complexity, which limit their clinical deployment. Developing efficient and lightweight networks is crucial for accurate lesion localization and optimized clinical workflows. We propose the VML-UNet, a lightweight segmentation network with core innovations including the CPMamba module and the multi-scale local supervision module (MLSM). The CPMamba module integrates the visual state space (VSS) block and a channel prior attention mechanism to enable efficient modeling of spatial relationships with linear computational complexity through dynamic channel-space weight allocation, while preserving channel feature integrity. The MLSM enhances local feature perception and reduces the inference burden. Comparative experiments were conducted on three public datasets, including ISIC2017, ISIC2018, and PH2, with ablation experiments performed on ISIC2017. VML-UNet achieves 0.53 M parameters, 2.18 MB memory usage, and 1.24 GFLOPs time complexity, with its performance on the datasets outperforming comparative networks, validating its effectiveness. This study provides valuable references for developing lightweight, high-performance skin lesion segmentation networks, advancing the field of skin lesion segmentation. Full article

To explore a more accurate prediction method for subgrade settlement induced by underpass construction, this study takes the existing railway project of Ningbo Yuanyi Road underpass as a case to construct a subgrade settlement prediction model based on the Mamba neural network. Using monitoring data collected using on-site automated monitoring robots as the data foundation, the prediction results of the improved transformer, long short-term memory (LSTM), time-series dense encoder (Tide), and decomposition-linear (Dlinear) neural networks are compared. The research results show that the Mean Squared Error (MSE) and Mean Absolute Error (MAE) of the proposed Bi-Mamba model are 0.279 and 0.276, respectively, demonstrating higher prediction accuracy than comparative models such as iTransformer and LSTM. Additionally, ablation experiments verify that the attention gating module in the model reduces the MSE by 9.1%, serving as a key component for improving accuracy. This study provides an advanced data-driven prediction method for subgrade settlement forecasting, offering technical references for similar engineering projects. Full article

Time series forecasting is critical for decision-making in numerous domains, yet achieving high accuracy across both short-term and long-term horizons remains challenging. In this paper, we propose a general hybrid forecasting framework that integrates a traditional statistical model (ARIMA) with modern deep learning models (such as LSTM and Transformer). The core of our approach is a novel multi-scale prediction mechanism that combines the strengths of both model types to better capture short-range patterns and long-range dependencies. We design a dual-stage forecasting process, where a classical time series component first models transparent linear trends and seasonal patterns, and a deep neural network then learns complex nonlinear residuals and long-term contexts. The two outputs are fused through an adaptive mechanism to produce the final prediction. We evaluate the proposed framework on eight public datasets (electricity, exchange rate, weather, traffic, illness, ETTh1/2, and ETTm1/2) covering diverse domains and scales. The experimental results show that our hybrid method consistently outperforms stand-alone models (ARIMA, LSTM, and Transformer) and recent, specialized forecasters (Informer and Autoformer) in both short-horizon and long-horizon forecasts. An ablation study further demonstrates the contribution of each module in the framework. The proposed approach not only achieves state-of-the-art accuracy across varied time series but also offers improved interpretability and robustness, suggesting a promising direction for combining statistical and deep learning techniques in time series forecasting. Full article

Using low-cost transition-metal chlorides and furfuryl alcohol as raw materials, the (TiZrHfNbTa)C precursor was prepared, and a three-dimensional braided carbon fiber preform (C/C) coated with pyrolytic carbon (PyC) was used as the reinforcing material. A C/C-(TiZrHfNbTa)C composite was successfully fabricated through the precursor impregnation pyrolysis (PIP) process. Under extreme oxyacetylene ablation conditions (2311 °C/60 s), this composite material demonstrated outstanding ablation resistance, with a mass ablation rate as low as 0.67 mg/s and a linear ablation rate of only 20 μm/s. This excellent performance can be attributed to the dense (HfZr)₆(TaNb)₂O₁₇ oxide layer formed during ablation. This oxide layer not only has an excellent anti-erosion capability but also effectively acts as an oxygen diffusion barrier, thereby significantly suppressing further ablation and oxidation within the matrix. This study provides an innovative strategy for the development of low-cost ultra-high-temperature ceramic precursors and opens up a feasible path for the efficient preparation of C/C-(TiZrHfNbTa)C composites. Full article

Spatiotemporal fusion techniques can generate remote sensing imagery with high spatial and temporal resolutions, thereby facilitating Earth observation. However, traditional methods are constrained by linear assumptions; generative adversarial networks suffer from mode collapse; convolutional neural networks struggle to capture global context; and Transformers are hard to scale due to quadratic computational complexity and high memory consumption. To address these challenges, this study introduces an end-to-end remote sensing image spatiotemporal fusion approach based on the Mamba architecture (Mamba-spatiotemporal fusion model, Mamba-STFM), marking the first application of Mamba in this domain and presenting a novel paradigm for spatiotemporal fusion model design. Mamba-STFM consists of a feature extraction encoder and a feature fusion decoder. At the core of the encoder is the visual state space-FuseCore-AttNet block (VSS-FCAN block), which deeply integrates linear complexity cross-scan global perception with a channel attention mechanism, significantly reducing quadratic-level computation and memory overhead while improving inference throughput through parallel scanning and kernel fusion techniques. The decoder’s core is the spatiotemporal mixture-of-experts fusion module (STF-MoE block), composed of our novel spatial expert and temporal expert modules. The spatial expert adaptively adjusts channel weights to optimize spatial feature representation, enabling precise alignment and fusion of multi-resolution images, while the temporal expert incorporates a temporal squeeze-and-excitation mechanism and selective state space model (SSM) techniques to efficiently capture short-range temporal dependencies, maintain linear sequence modeling complexity, and further enhance overall spatiotemporal fusion throughput. Extensive experiments on public datasets demonstrate that Mamba-STFM outperforms existing methods in fusion quality; ablation studies validate the effectiveness of each core module; and efficiency analyses and application comparisons further confirm the model’s superior performance. Full article

This paper proposes a comprehensive deep-learning framework, SentiStack, for Bitcoin price forecasting and trading strategy evaluation by integrating multimodal data sources, including market indicators, macroeconomic variables, and sentiment information extracted from financial news and social media. The model architecture is based on a Stacking-LSTM ensemble, which captures complex temporal dependencies and non-linear patterns in high-dimensional financial time series. To enhance predictive power, sentiment embeddings derived from full-text analysis using the DeepSeek language model are fused with traditional numerical features through early and late data fusion techniques. Empirical results demonstrate that the proposed model significantly outperforms baseline strategies, including Buy & Hold and Random Trading, in cumulative return and risk-adjusted performances. Feature ablation experiments further reveal the critical role of sentiment and macroeconomic inputs in improving forecasting accuracy. The sentiment-enhanced model also exhibits strong performance in identifying high-return market movements, suggesting its practical value for data-driven investment decision-making. Overall, this study highlights the importance of incorporating soft information, such as investor sentiment, alongside traditional quantitative features in financial forecasting models. Full article

Introduction: Percutaneous cryoablation (PCA) can be a valid alternative to partial nephrectomy for patients with cT1a renal tumors. A potential disadvantage of PCA is radiation exposure for patients, though the exact significance of this is unknown. This study aims to uncover the degree of radiation exposure during PCA and what factors are of influence. Methods: This is a retrospective analysis of a prospectively maintained database of patients who underwent CT-guided PCA for cT1 renal cell carcinoma (RCC) between January 2014 and September 2024. The median effective dose (mSV) of PCA was calculated and compared to the expected cumulative radiation exposure during follow-up. Multivariate linear regression was performed to identify factors predictive of higher radiation exposure (mSV). Results: A total of 164 PCAs were performed, with radiation data available for 133 cases. Mean age was 65 (±11) years and the mean tumor diameter was 28 (±9.6) mm. Median effective dose of the CA procedures was 26 mSV (IQR 18–37). The estimated cumulative effective dose of follow-up CT scans according to 2016 and 2024 European Association of Urology guidelines was 158 (IQR 117–213) and 105 mSV (IQR 78–142), respectively. Multivariate linear regression analysis identified BMI (OR 1.723, p < 0.001), the number of needles used (OR 4.060, p < 0.001), and the necessity for additional procedures (OR 8.056, p < 0.001) as significant predictors of a higher effective dose. Conclusions: We found a median effective dose of 26 mSV for PCA, which is relatively low compared to the cumulative radiation exposure associated with CT scans during follow-up of patients post-ablation according to the guidelines. Furthermore, increased BMI, a higher number of required needles and the execution of additional procedures are all associated with a higher effective dose. Full article

There is growing interest in measuring the temperature-dependent dielectric properties of bio-tissues using dual-mode techniques (scattering measurements and thermal treatment). Uncooled coaxial antennas are preferred for their direct contact with the measured medium and reduced complexity; however, they exhibit structural changes during ablation due to the thermal expansion of polytetrafluoroethylene (PTFE). This paper presents an experimental study on PTFE expansion in an uncooled coaxial insulated monopole antenna in response to changes in the tissue’s thermal environment. Furthermore, it presents a methodology to mitigate these effects through coaxial annealing. The investigation consists of two distinct experiments: characterising PTFE expansion and assessing the effects of annealing through microwave ablation. This was achieved by simulating the thermal effects experienced during ablation by immersing the test antenna in heated peanut oil. PTFE expansion was measured through camera monitoring and using a toolmaker’s microscope, revealing two expansion modalities: linear PTFE expansion and non-linear plastic deformation from manufacturing processes. The return loss during ablation and consequential changes in the ablated lesion were also assessed. Antenna pre-annealing increased resilience against structural changes in the antenna, improving lesion ellipticity. Therefore, this study establishes a fabrication method for achieving an uncooled thermally stable antenna, leading to an optimised dual-mode ablation procedure, enabling quasi-real-time permittivity measurement of the surrounding tissue. Full article

During rice cultivation, extracting levees helps to delineate effective planting areas, thereby enhancing the precision of management zones. This approach is crucial for devising more efficient water field management strategies and has significant implications for water-saving irrigation and fertilizer optimization in rice production. The uneven distribution and lack of standardization of levees pose significant challenges for their accurate extraction. However, recent advancements in remote sensing and deep learning technologies have provided viable solutions. In this study, Youyi Farm in Shuangyashan City, Heilongjiang Province, was chosen as the experimental site. We developed the SCA-UNet model by optimizing the UNet algorithm and enhancing its network architecture through the integration of the Convolutional Block Attention Module (CBAM) and Squeeze-and-Excitation Networks (SE). The SCA-UNet model leverages the channel attention strengths of SE while incorporating CBAM to emphasize spatial information. Through a dual-attention collaborative mechanism, the model achieves a synergistic perception of the linear features and boundary information of levees, thereby significantly improving the accuracy of levee extraction. The experimental results demonstrate that the proposed SCA-UNet model and its additional modules offer substantial performance advantages. Our algorithm outperforms existing methods in both computational efficiency and precision. Significance analysis revealed that our method achieved overall accuracy (OA) and F1-score values of 88.4% and 90.6%, respectively. These results validate the efficacy of the multimodal dataset in addressing the issue of ambiguous levee boundaries. Additionally, ablation experiments using 10-fold cross-validation confirmed the effectiveness of the proposed SCA-UNet method. This approach provides a robust technical solution for levee extraction and has the potential to significantly advance precision agriculture. Full article

In realistic hyperspectral image (HSI) cross-scene classification tasks, it is ideal to obtain target domain samples during the training phase. Therefore, a model needs to be trained on one or more source domains (SD) and achieve robust domain generalization (DG) performance on an unknown target domain (TD). Popular DG strategies constrain the model’s predictive behavior in synthetic space through deep, nonlinear source expansion, and an HSI generation model is usually adopted to enrich the diversity of training samples. However, recent studies have shown that the activation functions of neurons in a network exhibit asymmetry for different categories, which results in the learning of task-irrelevant features while attempting to learn task-related features (called “feature contamination”). For example, even if some intrinsic features of HSIs (lighting conditions, atmospheric environment, etc.) are irrelevant to the label, the neural network still tends to learn them, resulting in features that make the classification related to these spurious components. To alleviate this problem, this study replaces the common nonlinear generative network with a specific linear projection transformation, to reduce the number of neurons activated nonlinearly during training and alleviate the learning of contaminated features. Specifically, this study proposes a dimensionally decoupled spatial spectral linear extrapolation (SSLE) strategy to achieve sample augmentation. Inspired by the weakening effect of water vapor absorption and Rayleigh scattering on band reflectivity, we simulate a common spectral drift based on Markov random fields to achieve linear spectral augmentation. Further considering the common co-occurrence phenomenon of patch images in space, we design spatial weights combined with label determinism of the center pixel to construct linear spatial enhancement. Finally, to ensure the cognitive unity of the high-level features of the discriminator in the sample space, we use inter-class contrastive learning to align the back-end feature representation. Extensive experiments were conducted on four datasets, an ablation study showed the effectiveness of the proposed modules, and a comparative analysis with advanced DG algorithms showed the superiority of our model in the face of various spectral and category shifts. In particular, on the Houston18/Shanghai datasets, its overall accuracy was 0.51%/0.83% higher than the best results of the other methods, and its Kappa coefficient was 0.78%/2.07% higher, respectively. Full article

Thermal protection materials with excellent performance are critical for hypersonic vehicles. Carbon fiber/phenolic resin composites (C_f/Ph) have been widely used as thermal protection materials due to their high specific strength and ease of processing. However, oxidative failure limits the extensive applications of C_f/Ph in harsh environments. In this paper, a novel hafnium carbide (HfC) and boron carbide (B₄C)-modified C_f/Ph was fabricated via an impregnating and compression molding route. The synergistic effect of HfC and B₄C on the thermal stability, flexural strength, microstructure, and phase evolution of the ceramizable composite was studied. The resulting ceramizable composites exhibited excellent resistance to oxidative corrosion and ablation behavior. The residual yield at 1400 °C and the flexural strength after heat treatment at 1600 °C for 20 min were 46% and 54.65 MPa, respectively, with an increase of 79.59% in flexural strength compared to that of the composites without ceramizable fillers. The linear ablation rate (LAR) and mass ablation rate (MAR) under a heat flux density of 4.2 MW/m² for the 20 s were as low as −8.33 × 10⁻³ mm/s and 3.08 × 10⁻² g/s. The ablation mechanism was further revealed. A dense B–C–N–O–Hf ceramic layer was constructed in situ as an efficient thermal protection barrier, significantly reducing the corrosion of the carbon fibers. Full article

A triboelectric nanogenerator (TENG) holds significant potential as a self-powered pressure sensor due to its ability to convert mechanical energy into electrical energy. The output voltage of a TENG is directly correlated with the applied pressure, making it highly suitable for pressure sensing applications. Among the key factors influencing TENG performance, the microstructure on the surface plays a crucial role. However, the effect of surface microstructure on charge transfer behavior remains relatively underexplored. Here, a stripe-patterned rough TENG (SR-TENG) fabricated by laser ablation and molding is proposed. The stripe-patterned rough surface exhibits excellent deformation properties, allowing for more effective contact area between the tribolayers. Additionally, the localized surface-edge enhanced electric field at the stripe boundaries improves surface charge transfer, thereby enhancing overall output performance. The SR-TENG achieved an open-circuit voltage of 97 V, a short-circuit current of 59.6 μA, an instantaneous power of 3.55 mW, and a power density of 1.54 W/m². As an energy harvester, the SR-TENG successfully powered 150 LEDs. A linear relationship between applied pressure and output voltage was established with a coefficient of determination R² = 0.94, demonstrating a high sensitivity of 14.14 V/kPa. For practical application, a novel self-powered two-digit pressure switch was developed based on the SR-TENG. This system enables the control of two different LEDs using a single TENG device, triggered by applying a light or hard press. Full article