Search Results (1,138)

This study proposes a nondestructive method for evaluating egg freshness using laser Doppler vibrometry. A vibration measurement platform was constructed to collect vibration signals from intact eggs representing four freshness grades. The vibration signals were processed using multiplicative scatter correction, standard normal variate transformation, moving average filtering, Savitzky–Golay smoothing, and first-order Savitzky–Golay differentiation. Principal component analysis, successive projections algorithm, and competitive adaptive reweighted sampling were then used for feature extraction and dimensionality reduction. Six classification models were developed and compared, including support vector machine, k-nearest neighbor, random forest, naive Bayes, discriminant analysis, and linear discriminant analysis. The model based on moving average filtering, successive projections algorithm, and discriminant analysis using measurements from three egg locations achieved a classification accuracy of 96.7%. The model based on moving average filtering, competitive adaptive reweighted sampling, and random forest using blunt-end measurements also achieved an accuracy of 96.7%. These results demonstrate that laser Doppler vibrometry combined with machine learning can distinguish eggs with different freshness levels, providing a promising nondestructive approach for egg-quality evaluation. Full article

Accurate estimation of battery state indicators, particularly State of Charge (SoC) and State of Health (SoH), remains a fundamental requirement for advanced BMS, directly affecting safety, reliability, operational efficiency, and battery lifetime. Recent advances in deep learning have demonstrated promising results for SoC estimation and SoH classification individually; however, the impact of SoC estimation errors on subsequent SoH classification has received limited attention. This work investigates whether accurately estimated SoC information can effectively support data-driven SoH classification and quantifies the propagation of uncertainty across a cascaded battery diagnostic framework. A two-stage deep learning architecture combining LSTM-based SoC estimation and ResNet-based SoH classification is proposed and evaluated using three lithium-ion battery datasets. The SoC estimation stage is assessed using two experimental datasets (LG Chem E66 and Panasonic NCR18650PF), while the SoH classification and uncertainty-propagation analysis are conducted using the GS Yuasa LEV50 dataset generated from a validated battery simulation model. To improve temporal feature extraction and noise robustness, signal preprocessing based on Butterworth and Exponential Moving Average filters was incorporated. Experimental results achieved SoC estimation errors as low as 1.46% and an SoH classification accuracy of 94.23% when reference SoC values were used. Under realistic operating conditions, where SoH classification relied on estimated SoC values, the accuracy decreased to 79.65%. These findings provide a quantitative assessment of uncertainty propagation in integrated SoC-SoH pipelines and demonstrate that highly accurate SoC estimation alone does not necessarily guarantee equivalent SoH classification performance. The study highlights the challenges and limitations of cascaded data-driven battery diagnostics and provides insight for the development of more reliable intelligent BMS solutions. Full article

Off-grid photovoltaic (PV) hydrogen production systems must coordinate rapidly varying PV power, battery energy, and the operating states of multiple alkaline water electrolyzers. Inappropriate coordination may lead to PV curtailment, frequent unit switching, and persistent workload concentration on a small number of electrolyzers. This paper develops an efficiency- and load-balanced operation (ELBO) scheme as an improved rule-based supervisory strategy rather than an online optimization method. ELBO adopts a two-level decision structure. A planned number of online electrolyzers is first determined from the moving-average PV power and the reference power associated with high single-unit efficiency. This planned count is then corrected using real-time PV power, battery state of charge, and the previous electrolyzer states. The controller adjusts the powers of the online units before changing their number, uses the battery to bridge temporary power deficits, and distributes the remaining adjustable power under the operating and ramp-rate constraints. Five representative PV profiles selected from one year of measured data were used to compare ELBO with PV-following operation (PFO), multi-electrolyzer coordinated operation (MECO), and an offline mixed-integer linear programming (MILP) benchmark. ELBO produced 1328 kg of hydrogen, which was 8.85% and 6.07% higher than PFO and MECO, respectively. Its overall PV-to-hydrogen efficiency and PV utilization reached 65.2% and 94.9%, respectively, with 36 start–stop events. MILP produced 1345 kg of hydrogen, only 1.28% more than ELBO, but required the complete future PV sequence. Ablation analysis further shows that the planned-count layer, moving-average filtering, battery-supported retention, and load-balancing allocation contribute to different and complementary aspects of capacity matching, operating continuity, and workload distribution. The results indicate that the benefit of ELBO arises from the ordered coordination of these supervisory functions and that it provides a practical compromise between operating performance, workload distribution, information requirements, and computational complexity under the representative conditions considered. Full article

Toward low-carbon and sustainable power systems, hydro–wind–solar complementarity provides an important pathway for enhancing renewable energy accommodation and operational flexibility, while accurate characterization of uncertainty and cross-energy dependencies is essential for system optimization, operational risk assessment, and sustainable utilization of renewable resources. This study proposes a conditional Wasserstein generative adversarial network with gradient penalty integrating one-dimensional multi-scale channel attention (MSA) and an exponential moving average (EMA) mechanism (MSA-cWGAN-GP) for joint runoff–wind–photovoltaic (PV) scenario generation. The generator employs parallel depthwise 1D convolutions with multiple temporal receptive fields to capture multi-timescale variations, while an EMA shadow generator is used for model validation and scenario generation. Conditional labels are obtained by clustering joint 24 h runoff–wind–PV profiles, enabling generation under typical resource states. Case studies using historical runoff observations from Shuibuya Hydropower Station and wind and PV power series derived from ERA5 reanalysis data show overall absolute errors of the autocorrelation function (ACF) and Kendall coefficient of 0.0113 and 0.0495, respectively. The proposed model achieves the best average performance among the evaluated models in preserving intraday temporal dependence, cross-energy dependencies, and distributional characteristics, providing representative scenarios for uncertainty analysis and subsequent optimization of hydro–wind–solar complementary systems. Full article

Accurate forecasting of electricity prices is essential for efficient operation and decision-making in deregulated power markets, particularly in the Indian electricity market, characterized by high volatility and dynamic pricing. This study presents a comparative analysis of statistical, machine learning, and deep learning methods for electricity price prediction in the Indian market, with a focus on feature dimensionality. The evaluated models include autoregressive-integrated-moving average, seasonal autoregressive-integrated-moving average, seasonal autoregressive-integrated-moving average with exogenous variables, categorical boosting, random forest, long short-term memory, bidirectional long short-term memory, and a hybrid convolutional neural network-bidirectional long short-term memory model. Historical data available on the Indian Energy Exchange webpage are deployed in this study. The models are analyzed for varying input vector sizes with features that include date, type of day, day of the week, previous day, month, and year market prices. The results indicate improved predictive performance of all model while increasing the input feature dimensionality from five to seven. The results indicate that while increasing the feature size from five to seven increases the prediction accuracy, the gains become marginal beyond seven, emphasizing the importance of feature relevance over feature quantity. From a theoretical perspective, the study highlights the dominance of short-term temporal dependencies in MCP prediction and provides empirical evidence for the point of diminishing returns in feature expansion. From a practical standpoint, the results endorse the choice of computationally efficient and interpretable models for real-world deployment. Full article

Decarbonizing the transportation sector depends not only on the scale of mitigation programs but also on whether financing systems are capable of generating measurable emission reductions. In Indonesia, climate finance allocation remains substantially below the level required to achieve the transportation-sector target under the Enhanced Nationally Determined Contribution (ENDC). At the same time, mitigation planning rarely establishes a clear relationship between financial expenditure and verified greenhouse gas (GHG) reduction outcomes, making policy effectiveness difficult to assess. This study examines the relationship between climate finance and mitigation outcomes in Indonesia’s transportation sector using verified emission reduction data and realized mitigation expenditures during 2018–2022. A cost-based assessment approach was applied to estimate the financing required to reduce one ton of CO₂ equivalent (tCO₂-e) across direct and indirect mitigation actions. The analysis identified 33 mitigation actions categorized under the Avoid–Shift–Improve (ASI) framework and evaluated their contribution to sectoral emission reduction. The results indicate substantial variation in mitigation costs among intervention types. Direct mitigation actions, particularly mass public transportation expansion, are linked to larger emission reductions at relatively lower costs than enabling or indirect measures. On average, reducing 1 tCO₂-e in Indonesia’s transportation sector requires approximately USD 184–1000 (IDR 3–16.4 million), using a standardized exchange rate of approximately IDR 16,400 per USD. Based on the transportation-sector ENDC target, the estimated financing requirement by 2030 ranges from USD 2.5–13.8 billion (IDR 42–226 trillion). The findings suggest that climate finance policies should move beyond expenditure-oriented approaches toward financing frameworks that explicitly connect investment allocation with verified mitigation performance. Full article

Background: Environmental exposures are biologically plausible contributors to thyroid cancer outcomes, but population-based evidence is heterogeneous and often focuses on incidence rather than stage at diagnosis. Objective: This study aimed to evaluate whether ecological state-year environmental indicators add predictive information for advanced-stage thyroid cancer at diagnosis in a large SEER-based cohort (U.S. National Cancer Institute’s Surveillance, Epidemiology, and End Results Program). Methods: We conducted a retrospective registry-linked analysis of 368,726 SEER thyroid cancer cases with valid combined summary stage. Advanced stage was defined as regional or distant disease and occurred in 118,509 cases (32.14%). Environmental variables were linked as state-year indicators using five-year moving averages from the years preceding diagnosis. Regularized logistic regression, gradient boosting, and Extra Trees models evaluated incremental predictive performance. Results: Adding environmental variables to demographic, socioeconomic, histology, and laterality predictors produced small random-split improvements across model families. However, after SEER registry was added before environmental variables, the remaining environmental increment was minimal. Registry-group holdout validation did not support improved geographic generalizability from the environmental feature set. Conclusions: State-year environmental indicators carried limited predictive information for advanced thyroid cancer stage, and much of this information overlapped with registry and geographic structure. These findings clarify both the potential and limitations of ecological environmental linkage in SEER-based prediction studies. Full article

Expectations regarding collective climate action are multidimensional and should be understood in relation to the social contexts in which they are embedded. This study examines whether these expectations are associated with the country-level average belief in the anthropogenic origin of climate change and with individuals’ deviation from that average. Drawing on Round 10 of the European Social Survey, the analysis utilizes 24,014 valid cases from 29 European countries. It estimates Complex Samples General Linear Models (CSGLM) to account for the complex sampling design. The results show that the country-level average belief in the anthropogenic origin of climate change is positively associated with expectations of collective climate action; that is, countries with a higher average belief exhibit higher expectations of collective climate action. Conversely, individuals whose belief in the anthropogenic origin of climate change exceeds the respective country-level average tend to express lower expectations of collective climate action. This negative association becomes stronger at higher levels of climate worry and is attenuated as ideological positioning moves towards the right; in contrast, trust in scientists does not have a significant moderating effect. Overall, the findings support a socially contextualized understanding of expectations of collective climate action, in which both national belief contexts and individuals’ relative positioning within those contexts are relevant. Full article

Environmental noise is a major environmental and public health concern in rapidly urbanizing cities, yet integrated studies combining field measurements, GIS-based spatial analysis, and predictive modelling remain limited in Central Asia. This study assessed the spatial distribution of urban noise pollution in Samarkand (Uzbekistan) and explored its statistical association with selected public health indicators, forecasting future trends. Measurements were conducted at 50 georeferenced sites covering more than 300 streets. The measured data were processed and mapped using ArcGIS 10.5 (Esri, Redlands, CA, USA) to produce the spatial distribution of environmental noise across the study area. Official data on registered vehicles, industrial enterprises, and disease incidence (2014–2024) were analysed using Pearson correlation, Autoregressive Integrated Moving Average (ARIMA), and its extension incorporating exogenous variables (ARIMAX) models. Results revealed pronounced spatial heterogeneity in noise levels, highest along transport corridors and industrial zones. Industrial enterprises showed the strongest correlations with disease incidence; vehicle registrations were excluded from final models owing to collinearity with industrial activity. ARIMA projected continued industrial growth through 2030, while ARIMAX models identified significant associations between industrial activity and diseases of the ear and mastoid process and of the nervous system (MAPE 28.55% and 19.78%). As an ecological, exploratory study using infrastructural proxies rather than measured noise exposure, findings should be interpreted as associations rather than causation. The framework offers a transferable approach for environmental risk assessment and sustainable urban planning. Full article

This study develops a rigorous, leakage-free forecasting framework for monthly Robusta coffee prices using historical observations from January 1975 to December 2025. A comprehensive set of explanatory variables is constructed from lagged coffee prices, moving averages, logarithmic returns, rolling volatility, and exogenous variables such as the Oceanic Niño Index (ONI), the U.S. Dollar Index, and Brent crude oil prices. To ensure methodological fairness, all predictors are generated exclusively from information available at the forecast origin, and all competing models are evaluated under a unified expanding-window walk-forward validation framework. Seven forecasting models are compared: Naïve, Exponential Smoothing (ETS), ARIMA, ARIMAX, Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). Forecasting performance is evaluated using R², RMSE, MAE, and MAPE, while Taylor diagrams and the Diebold–Mariano test are employed to assess model agreement and differences in predictive accuracy. The results show that XGBoost achieves the highest forecasting accuracy (R² = 0.956, RMSE = 0.264), followed closely by the Naïve (R² = 0.954, RMSE = 0.271) and ARIMA (R² = 0.954, RMSE = 0.270) benchmarks, whereas ARIMAX and ETS provide comparable performance and the deep learning models (LSTM and GRU) produce substantially larger prediction errors. Feature importance analysis further indicates that the first lag of coffee price is the dominant predictor, accounting for approximately 94% of the predictive gain in XGBoost. Overall, the findings demonstrate that rigorous leakage-free validation is essential for reliable forecasting research and that, for monthly Robusta coffee prices, increased model complexity does not necessarily yield superior predictive performance. Full article

Fine grinding is often associated with high energy consumption, and the choice of operating parameters has a direct effect on product fineness. This study examines dry particle size reduction under the two-sided impact of counter-moving grinding media. A previously developed calculation model was used to estimate the final particle size, and the calculated values were compared with experimental results obtained for quartz-rich river sand with an initial particle size of 100 μm. The study considered four parameters: horizontal amplitude, vertical amplitude, oscillation frequency, and grinding media diameter. Increasing the horizontal amplitude from 2 to 6 mm reduced the experimental product fineness from 19.6 to 4.8 μm. The vertical amplitude had a weaker effect, but increasing it from 1 to 4 mm also reduced the experimental particle size from 9.9 to 5.0 μm. An increase in frequency from 40 to 60 Hz reduced the experimental product fineness from 12.1 to 5.4 μm. The opposite tendency was observed for grinding media diameter: increasing it from 5 to 15 mm increased the experimental particle size from 4.8 to 11.4 μm. The average difference between theoretical and experimental results was 12.28%. In 12 of the 17 studied operating modes, the deviation was less than 15%. Laser particle size analysis of the selected finest sample was additionally performed to characterize the particle size distribution and fine particle fraction. The present work extends the previous model validation by evaluating the main operating variables and proposing a preliminary rational operating range for the studied dry grinding process. The results show that the model can be used for preliminary estimation of product fineness and for identifying the most influential operating parameters under the studied dry grinding conditions. Full article

Top predators drive ecosystem functions. Reintroductions of top predators are a vital conservation management strategy to achieve this. Ten spotted hyenas (hyenas—Crocuta crocuta) were reintroduced to the Tswalu Kalahari Reserve (TKR). To determine how hyenas recolonised the TKR, we fitted five hyenas with satellite collars. We assessed home range development and stabilisation using Movement Kernel Density Estimation (MKDE), home range overlap and dynamic interactions between individuals using proximity analysis. From hourly locations, we compared movement rates between individuals. We found that hyena home ranges stabilised within seven months of release. The average home range for reintroduced hyenas was 307 km² (SD = 124 km²). Home ranges overlapped ≤36% between individuals. Hyenas’ movements synchronised with each other at distances ≤ 10 km. The hyenas moved farther hourly in cooler seasons, and they displayed unimodal activity patterns (longest hourly movements in the afternoon and night). We found that hyenas reintroduced into the TKR (without other top predators) required <7 months to re-establish themselves, and their social dynamics are not strictly tied to those individuals with which they were reintroduced. In arid settings where there are no lions and resources are abundant, home ranges for reintroduced hyenas may be smaller than is typical for this species. Full article

Accurately modeling the auxiliary power consumption of Battery Energy Storage Systems (BESSs) is increasingly important as grid-scale storage assets are becoming involved in electricity markets. In this paper, we develop a data-driven framework to characterize and forecast auxiliary consumption using operational data from a utility-scale BESS deployed in Italy. To capture the short-term thermal inertia of the system and the delayed response of the cooling systems, a set of predictors based on moving averages of power and ambient temperature is constructed. Two novel modeling approaches are proposed: a three-dimensional look-up table (LUT) representation that provides an interpretable characterization of system behavior, as well as a Random Forest (RF) regression model capable of capturing complex non-linear relationships between parameters. These are evaluated in comparison with a two-dimensional LUT from the literature. The analysis showed a superior performance of the RF model that comes at the cost of reduced interpretability and computational efficiency, while both proposed models outperform the literature-based LUT. Moreover, the impact of the number and type of predictors on model performance is systematically assessed, to shed light on what constitutes the requirements for a reasonably accurate estimation of the auxiliary systems. Finally, the concept of forecasting uncertainty for the temperature and day-ahead forecasting applicability for the auxiliary systems is investigated by introducing a persistence-based logic and an evaluation of the impact on the results. Overall, the study highlights the importance of explicitly modeling auxiliary consumption in grid-scale BESSs, proposes well-performing models for its estimation, and provides practical guidelines for their implementation in energy management and forecasting applications. Full article

Reliable visibility forecasting is important for transportation safety and sustainable urban environmental management, particularly in the Sichuan Basin, where poor visibility remains a persistent concern. This study developed a comparative diagnostic framework for one-day-ahead prediction of continuous daily visibility in Chengdu using meteorological and air-quality observations from November 2021 to October 2024. Grey relational analysis and temporal diagnostics were used to characterize variable associations and temporal dependence. Seven methods—seasonal autoregressive integrated moving average with exogenous variables (SARIMAX), CatBoost, long short-term memory (LSTM), Transformer, convolutional neural network–LSTM (CNN–LSTM), CNN–Transformer, and Transformer–LSTM—were evaluated under four input configurations, yielding 28 model–input combinations. SARIMAX achieved the best overall balance between predictive accuracy and generalization stability. CatBoost was the most accurate data-driven method and obtained the highest coefficient of determination (R² = 0.644) under meteorological-only inputs. Meteorological-only inputs outperformed pollutant-only and full multivariate inputs for all models. Among the deep-learning models, LSTM performed better under univariate, meteorological-only, and pollutant-only inputs, whereas Transformer benefited more from the full multivariate input. CNN preprocessing improved LSTM mainly under the full multivariate and pollutant-only configurations. The moderate maximum R² suggests that routine observations did not fully capture aerosol composition, particle number and size distributions, hygroscopic growth, aerosol–water interactions, and fog-related processes. These findings support model and input selection for daily visibility forecasting. Full article

To improve the accuracy and stability of new energy vehicle (NEV) sales forecasting in China, this study develops a hybrid forecasting framework integrating the Informer model, autoregressive integrated moving average (ARIMA), and the Boomerang Aerodynamic Ellipse Optimization (BAEO) algorithm. Monthly data from January 2016 to December 2023 covering 31 provincial-level administrative regions in China (excluding Hong Kong, Macao, and Taiwan) were collected from authoritative statistical sources. A multidimensional feature system was established by incorporating factors related to charging infrastructure, transportation demand, market development, and environmental conditions. Data preprocessing techniques, including Min–Max normalization, lagged variables, rolling statistical features, and seasonal sine–cosine encoding, were applied to capture temporal dependencies and periodic patterns. The BAEO algorithm was employed to optimize the key hyperparameters of the Informer model, while the ARIMA model was introduced to correct linear patterns in forecasting residuals. The proposed BAEO–Informer–ARIMA framework was evaluated against seasonal autoregressive integrated moving average (SARIMA), Prophet, extreme gradient boosting (XGBoost), long short-term memory (LSTM), gated recurrent unit (GRU), Transformer, and Informer models under the same chronological evaluation strategy. Results show that the proposed framework achieved superior forecasting performance, with a coefficient of determination (R²) of 0.9544, mean absolute error (MAE) of 24,068, root mean square error (RMSE) of 26,822, and mean absolute percentage error (MAPE) of 3.39%. Furthermore, uncertainty analysis based on rolling-validation forecast errors was conducted to establish a 90% confidence interval for future projections. Forecast results for 2024–2030 reveal sustained NEV sales growth with gradually decreasing growth rates and persistent seasonal variations. This study provides quantitative insights for NEV market planning, charging infrastructure deployment, and low-carbon policy formulation. Full article