Optimal Model Research Articles

Optimal scheduling of a cogeneration system via Q-learning-based memetic algorithm considering demand-side response

An optimization scheduling method for cogeneration systems based on the Q-learning-based memetic algorithm (QMA) is developed to enhance the system's wind power absorption capacity and economical operation. First, an optimal scheduling model for cogeneration systems is constructed. The model integrates time-of-use electricity pricing and heating comfort into the demand side, fully exploiting the potential of demand-side response (DR) in power generation and heating. An additional heat source (AHS), which consists of an electric boiler and heat storage tank, is used to uncouple heat and power. In addition, a QMA algorithm is designed to find the optimal operating scheme for the cogeneration system. The Q-learning algorithm is introduced to dynamically adjust crossover and mutation parameters during the global evolution stage, improving the algorithm's search capability. The Taguchi method is utilized for algorithm parameter calibration. Finally, the simulation results under various operating scenarios are compared and analyzed, verifying the feasibility and effectiveness of the proposed method. The simulation results show that DR and AHS can improve the system's economic performance and wind power utilization rate. Compared with the IMA, IPSO, MA, and IABC algorithms, the QMA algorithm reduces the average economic cost by approximately 1.18%, 2.23%, 5.69%, and 14.69%, respectively.

Durability of concrete-encapsulated GFRP bars subjected to seawater immersion

Glass fiber reinforce polymer (GFRP) bars have better durability than traditional steel bars, but there are few studies on their long-term performance in actual service environments especially the marine environment. To investigate the durability of GFRP bars used in cross-sea bridges, this research delves into the performance of 200 such bars, spanning diameters of 10, 12, 16, and 25mm, in an environment that closely mimics real-world conditions. The GFRP bars, encapsulated in concrete, were submerged in seawater solutions, maintained at either room temperature or 60°C, for periods spanning from 0 to 183 days. Then the tensile strength and elastic modulus degradation of GFRP bars were investigated by tensile tests. Scanning electron microscope (SEM) was also used to analyze the micro-degradation mechanism of GFRP bars. In addition, this paper compares the four existing models for predicting long-term mechanical properties of GFRP bars with the experimental data, so as to determine the formulas and methods to be used in the design. The result shows that the elastic modulus of GFRP bars after 183 days of immersion in concrete pore solution was almost unchanged while the tensile strength decreased significantly. Among them, the elastic modulus and tensile strength of 10mm diameter GFRP bars decreased by 5% and 35%, respectively, after 183 days of immersion at high temperatures (60°C). The SEM method revealed that the main cause of such phenomena was the weakening of the bond between the fibers and the matrix, which led to a decrease in the interface properties. Finally, by fitting and comparing the relationship between the test data and the existing prediction models for the long-term mechanical properties of GFRP bars, the residual tensile strength of GFRP bars used in the actual project of Yanzhou Bridge for 100 years was predicted according to the optimal prediction model, i.e. the Bank model, and the results showed that the residual strength of GFRP bars is able to meet the current standard requirements.

Effects of various cold plasma atmospheres and external influencing factors on the adsorption of formaldehyde gas by bamboo-based carbon microfibers

To effectively enhance the adsorption performance of gaseous formaldehyde (HCHO) by bamboo-based carbon microfibers (BCMFs), the active oxygen (O) and nitrogen (N)-containing functional groups were rapidly introduced by cold plasma. The effects of plasma time, plasma atmospheres including oxygen (O2), nitrogen (N2), air, and argon (Ar) as well as ambient temperature, flow rate, and initial concentration of gaseous HCHO on the adsorption performance of BCMFs were comprehensively investigated. The typical kinetic, isothermal, and thermodynamic models were also evaluated. The O2-plasma activated BCMFs showed the most serious etching effects among four atmospheres. The adsorption capacities were ranked as BCMF-800-N-10 (203.87 mg g−1) > BCMF-800-Ar-10 (166.26 mg g−1) > BCMF-800-Air-10 (151.28 mg g−1) > BCMF-800-O-10 (144.21 mg g−1), which were 112.99%, 73.69%, 58.04%, and 50.66% higher than the control, respectively. The O–CO and pyrrolic N groups were more easily formed onto BCMF-800-N-10. The Langmuir and the Freundlich models displayed the excellent fitting properties. The optimal kinetic model tended to transfer to the pseudo-second-order from the pseudo-first-order after cold plasma modification. The BCMFs activated by plasma showed a synergistic single/multi-layer adsorption mechanism with a complex physicochemical adsorption. The HCHO adsorption performance was positively correlated with its concentrations but negatively correlated with environmental temperature.

The prognostic role of an optimal machine learning model based on clinical available indicators in HCC patients

The prognostic role of an optimal machine learning model based on clinical available indicators in HCC patients

Machine-learning-aided regional post-seismic usability prediction of buildings: 2016-2017 Central Italy earthquakes

This study addresses the development of machine learning models for predicting the post-seismic buildings usability within regions prone to frequent earthquakes. The analysis leverages on field data from 2016-2017 Central Italy earthquakes. Several machine learning techniques are employed for this task, namely K-nearest neighbors, linear support vector machine, radial basis function support vector machine, decision tree, random forest, neural network, adaptive boosting, naive Bayes, quadratic discriminant analysis, logistic regression, and linear discriminant analysis. The input variables include both building attributes and seismic intensity measures. Since the database turns out to be strongly imbalanced, the potential influence of two preprocessing techniques is examined, namely principal component analysis and synthetic minority oversampling technique. Several metrics are considered to evaluate the performance of the resulting predictive machine learning models. Moreover, this study investigates the optimal machine learning model’s robustness against uncertainties, quantifies the importance of its features, and investigates how usability classes clustering can impact its performance. Every step of the implemented procedure is deeply explained and discussed to provide useful guidelines for similar applications.

Hybrid Scheduling Strategy and Improved Marine Predator Optimizer for Energy Scheduling in Integrated Energy System to Enhance Economic and Environmental Protection Capability

This study endeavors to comprehensively enhance energy efficiency and investigate the potential of integrated energy system in energy conservation and emission mitigation by proposing innovative solutions. Consequently, an optimal scheduling model has been developed. The analytic hierarchy process is employed to ascertain the weights of indicators within the model. Strategies for enhanced following heating load and enhanced following power load have been introduced, considering the energy transfer capability of the energy storage system. Furthermore, a hybrid strategy enabling flexible interchange between these two strategies has been proposed. Subsequently, an improved marine predator optimizer is proposed to solve the model, which significantly boosts solution efficiency and the capability to overcome local optima. Results indicate an overall enhancement in system performance under the proposed strategies compared to traditional approaches. The hybrid strategy particularly stands out, demonstrating superior performance with energy efficiency achieving up to 75.67% and the comprehensive performance improvement of 5.01% and 7.68% compared to traditional following power load strategy and following heating load strategy. This study contributes to reducing system operating costs, energy consumption, and carbon emissions, signifying its substantial role in advancing renewable energy integration, promoting green economies, and fostering sustainable development.

Machine Learning-Based Radiomics for Predicting Outcomes in Cervical Cancer Patients Undergoing Concurrent Chemoradiotherapy

Purposes To investigate the value of machine learning-based radiomics for predicting disease-free survival (DFS) and overall survival (OS) undergoing concurrent chemoradiotherapy (CCRT) for patients with locally advanced cervical cancer (LACC). Materials and methodsIn this multicentre study, 700 patients with IB2-IVA cervical cancer who underwent CCRT with ongoing follow-up were retrospectively analyzed. Three-dimensional radiomics features of primary lesions and its surrounding 5 cm region in T2WI sequences were collected. Six machine learning methods were used to construct the optimal radiomics model for accurate prediction of DFS and OS after CCRT in LACC patients. Eventually, TCGA and GEO databases were used to explore the mechanisms of radiomics in predicting the progression and survival of cervical cancer. This study adhered CLEAR for reporting and its quality was assessed using RQS and METRICS. ResultsIn the prediction of DFS, the RSF model combined tumor and peritumor radiomics demonstrated the best predictive efficacy, with the AUC for predicting 1-year, 3-year, and 5-year DFS in the training, validation, and test sets of 0.986, 0.989, 0.990, and 0.884, 0.838, 0.823, and 0.829, 0.809, 0.841, respectively. In the prediction of OS, the GBM model best performer, with AUC of 0.999, 0.995, 0.978, and 0.981, 0.975, 0.837, and 0.904, 0.860, and 0.905. Differential genes in TCGA and GEO suggest that the prediction of radiomics model may be associated with KDELR2 and HK2. ConclusionMachine learning-based radiomics models help to predict DFS and OS after CCRT in LACC patients, and the combination of tumor and peritumor information has higher predictive efficacy, which can provide a reliable basis for therapeutic decision-making in cervical cancer patients.

Changes in Characteristics of Spermatogonial Stem Cells in Response to Heat Stress in Stallions

Spermatogonial stem cells (SSCs) are essential for the maintenance of male fertility and survival of species. Environmental conditions, notably heat stress, have been identified as important causes of male infertility and have a negative impact on SSCs. Animals with cryptorchid testes (CT) are optimal models for the study of long-term heat stress-related changes in germ cells. The effect of heat stress on germ cells differs depending on the spermatogenesis stage. Thus, verifying whether the specific phase of spermatogenesis is dependent or independent of heat stress in stallions is important. We evaluated the heat stress-related response of SSCs by comparing the relative abundance of mRNA transcripts and expression patterns of the undifferentiated embryonic cell transcription factor 1 (UTF-1) and deleted in azoospermia-like (DAZL) in the seminiferous tubules of CT and normal testes (NT) of stallions using reverse transcription-quantitative polymerase chain reaction (RT-qPCR), immunofluorescence, and western blotting. We also analyzed the relative abundance of mRNA of different proliferative markers, including minichromosome maintenance 2 (MCM2), marker of proliferation Ki-67 (MKI-67), and proliferating cell nuclear antigen (PCNA). Testicular tissues from four Thoroughbred unilateral cryptorchid postpubertal stallions were used in this study during the breeding season. The relative abundance of the mRNA transcripts of UTF-1 and MCM2 was significantly upregulated in the CT group than that of those in the NT group. In contrast, the relative abundance of the mRNA transcripts of DAZL was significantly downregulated in the CT group than that of those in the NT group. Western blot quantification showed that the relative intensity of UTF-1 protein bands was significantly higher, while that of DAZL protein bands was significantly lower in the CT group than in the NT group. Immunofluorescence studies showed that the number of germ cells immunostained with UTF-1 was significantly higher while immunostained with DAZL was significantly lower in the CT group than that in the NT group. The higher expression level of UTF-1 in the CT group shows that undifferentiated SSCs are not affected by long-term exposure to heat stress. These results also indicate that germ cells after differentiation phase are directly affected by heat-stress conditions, such as cryptorchidism, in stallions.

Machine learning in epidemiology: Neural networks forecasting of monkeypox cases.

This study integrates advanced machine learning techniques, namely Artificial Neural Networks, Long Short-Term Memory, and Gated Recurrent Unit models, to forecast monkeypox outbreaks in Canada, Spain, the USA, and Portugal. The research focuses on the effectiveness of these models in predicting the spread and severity of cases using data from June 3 to December 31, 2022, and evaluates them against test data from January 1 to February 7, 2023. The study highlights the potential of neural networks in epidemiology, especially concerning recent monkeypox outbreaks. It provides a comparative analysis of the models, emphasizing their capabilities in public health strategies. The research identifies optimal model configurations and underscores the efficiency of the Levenberg-Marquardt algorithm in training. The findings suggest that ANN models, particularly those with optimized Root Mean Squared Error, Mean Absolute Percentage Error, and the Coefficient of Determination values, are effective in infectious disease forecasting and can significantly enhance public health responses.

Enhancing the exploitation of natural resources for green energy: An application of LSTM-based meta-model for aluminum prices forecasting

Efficient resource allocation for electric car production can be achieved by anticipating aluminum future pricing. For the electric car sector to maintain a steady supply chain, efficient production, and long-term growth, accurate forecasts are essential for policymakers to plan resource policies. The accuracy of ensemble learning models in predicting Aluminum prices was analyzed using Gene Expression programming and long short-term memory (LSTM) method. The analysis was based on monthly frequency spot settlement price data of Aluminum, copper, silver, and crude oil prices, as well as currency inflation rates of USA, China, and Peru. The study period spanned from January 1, 1994 to December 31, 2022, and also included the gross domestic product (GDP) of USA, China, and other metal minerals. Initially, we assess the level of multicollinearity between each chosen input parameter and Aluminum price volatility using Variance inflation factor-based multicollinearity. The primary innovation of this study lies in the creation of a highly precise model that utilizes the LSTM algorithm to estimate monthly Aluminum prices. This model serves as a revolutionary forecasting method. An extensive evaluation of various ensemble learning models reveals that XGBoost is the optimal model for accurately forecasting monthly Aluminum prices. The LSTM-based recurrent neural network model exhibited the lowest error, with a Mean Relative Error (MRE) ranging from 0.001 to 0.008, RMSE ranging from 54.527 to 136.044, and R2 ranging from 0.95 to 0.983. The findings demonstrated that the LSTM-based super learner models had superior performance compared to the CatBoost, LightBoost, Random Forest, XGBoost, and AdaBoost models, as evidenced by higher values of the determination coefficient (R2), value account for (VAF), and other 10 error analysis evaluators. Therefore, the utilization of an LSTM-based super learner model can serve as a dependable approach for forecasting future monthly Aluminum prices. In addition, this study introduces a logical mathematical model utilizing gene expression programming (GEP) to anticipate future Aluminum prices, which can be utilized by other researchers. Thus, we conclude that both the GEP and ensemble learning models, particularly the LSTM-based Super learner model, are appropriate for precise metal price predictions, which can benefit policymakers and assist in resource policy planning.

The Intuitive Eating Scale-3: Development and psychometric evaluation

Intuitive eating has been found to protect against disordered eating and preserve well-being. Yet, there are methodological (length), conceptual (inconsideration of medical, value-based, and access-related reasons for food consumption), and psychometric (item wording) concerns with its most common measure, the Intuitive Eating Scale-2 (IES-2). To address these concerns, we developed the IES-3 and investigated its psychometric properties with U.S. community adults. Across three online studies, we evaluated the IES-3's factorial validity using exploratory factor analysis (EFA; Study 1; N = 957; Mage = 36.30), as well as confirmatory factor analysis (CFA), exploratory structural equation modeling (ESEM), bifactor-CFA, and bifactor-ESEM (Study 2; N = 1152; Mage = 40.95), and cross-validated the optimal model (Study 3; N = 884; Mage = 38.54). We examined measurement invariance across samples and time, differential item functioning (age, body mass index [BMI], gender), composite reliability, and validity. Study 1 revealed a 12-item, 4-factor structure (unconditional permission to eat, eating for physical reasons, reliance on hunger and satiety cues, body-food choice congruence). In Study 2, a bifactor-ESEM model with a global intuitive eating factor and four specific factors best fit the data, which was temporally stable across three weeks. This model also had good fit in Study 3 and, across Studies 2 and 3, and was fully invariant and lacked measurement bias in terms of age, gender, and BMI. Associations between latent IES-3 factors and age, gender, and BMI were invariant across Studies 2 and 3. Composite reliability and validity (relationships with disordered eating, embodiment, body image, well-being, and distress; negligible relationship with impression management) of the retained model were also supported. The 12-item IES-3 demonstrates strong psychometric properties in U.S. community adults. Research is now needed using the IES-3 in other cultural contexts and social identity groups.

Mechanical Behavior of Graded Combined Auxetic-Honeycomb Structures

Various methods such as geometry changes can improve the energy absorption of auxetic structures. In this article, the mechanical behavior of graded combined auxetic-honeycomb structures is investigated through various models of grading. In the graded angle method, the best samples are selected for construction and experimental after performing numerical simulations, and this method comprises both the all-auxetic and honeycomb-auxetic graded angle techniques. Additionally, it is proven that energy absorption can be further increased by adding a reinforced core to the cells of the re-entrant structure. Finally, optimal models for energy absorption are developed by combining the best models of graded angle, graded thickness, and reinforced core. The best of them is the all-auxetic unilateral graded angle with rhombic core auxetic (UGA-A)+(RCA) sample with energy absorption that is three times higher than HAR, which can be utilized in various industrial and sports applications.

Temporal Forecast of Maize Production in Tanzania: An Autoregressive Integrated Moving Average Approach

This paper aims to model and forecast maize production in Tanzania, emphasizing its crucial contribution to the agricultural sector. Utilizing secondary data on annual maize production in Tanzania from 1961 to 2021, measured in tonnes and sourced from the FAO database, the study employs statistical techniques such as unit root tests, the ARIMA model, and the Ljung-Box test for a comprehensive analysis. The optimal model for forecasting maize production in Tanzania is identified as ARIMA (5,1,5), and the adequacy assessment confirms its effectiveness in predicting maize production values. The ten-year forecast reveals an intermittent pattern, offering valuable insights into anticipated trends. While not explicitly developing an existing theory, this research enhances the understanding of maize production dynamics in Tanzania through the application of advanced statistical methods. Policymakers, agricultural stakeholders, and researchers can leverage these findings to enhance agricultural productivity and food security in Tanzania.

Evaluation of transformation invariant loss function with distance equilibrium in prediction of imaging photoplethysmography characteristics

Objective. Monitoring changes in human heart rate variability (HRV) holds significant importance for protecting life and health. Studies have shown that Imaging Photoplethysmography (IPPG) based on ordinary color cameras can detect the color change of the skin pixel caused by cardiopulmonary system. Most researchers employed deep learning IPPG algorithms to extract the blood volume pulse (BVP) signal, analyzing it predominantly through the heart rate (HR). However, this approach often overlooks the inherent intricate time-frequency domain characteristics in the BVP signal, which cannot be comprehensively deduced solely from HR. The analysis of HRV metrics through the BVP signal is imperative. Approach. In this paper, the transformation invariant loss function with distance equilibrium (TIDLE) loss function is applied to IPPG for the first time, and the details of BVP signal can be recovered better. In detail, TIDLE is tested in four commonly used IPPG deep learning models, which are DeepPhys, EfficientPhys, Physnet and TS_CAN, and compared with other three loss functions, which are mean absolute error (MAE), mean square error (MSE), Neg Pearson Coefficient correlation (NPCC). Main results. The experiments demonstrate that MAE and MSE exhibit suboptimal performance in predicting LF/HF across the four models, achieving the Statistic of Mean Absolute Error (MAES) of 25.94% and 34.05%, respectively. In contrast, NPCC and TIDLE yielded more favorable results at 13.51% and 11.35%, respectively. Taking into consideration the morphological characteristics of the BVP signal, on the two optimal models for predicting HRV metrics, namely DeepPhys and TS_CAN, the Pearson coefficients for the BVP signals predicted by TIDLE in comparison to the gold-standard BVP signals achieved values of 0.627 and 0.605, respectively. In contrast, the results based on NPCC were notably lower, at only 0.545 and 0.533, respectively. Significance. This paper contributes significantly to the effective restoration of the morphology and frequency domain characteristics of the BVP signal.

Surrogate-based Shape Optimization Design for the Stable Descent of Mars Parachutes

The crucial role that the supersonic parachute plays in space exploration missions has been widely recognized, as it directly impacts the safe landing of probes. However, parachute models with optimization on different aerodynamic performances often involve design conflicts with each other. Additionally, the parachute design focusing on a single point cannot fully adapt to different speed ranges during stable descent. This complexity makes it challenging to use traditional shape design methods, which rely on empirical knowledge, to address these coupled design issues. Faced with the design challenges of Mars parachutes, this study, inspired by aircraft aerodynamic optimization principles, establishes a shape design method specifically for the stable descent phase of Mars parachutes. The method combines numerical simulation and surrogate-based optimization strategies, aiming to enhance the overall performance during stable descent and meet various demands of different exploration missions. Meanwhile, by providing a rapid estimate of the shape during the design phase, the method significantly improves computational efficiency. The optimal models effectively balance comprehensive performance in the supersonic-transonic-subsonic speed domain by conducting shape optimization research on the disk-gap-band parachute using the surrogate-based optimization strategy. Also, it exhibits better deceleration and stability across the entire speed range compared to the base model, even when deviating from the design Mach number. Importantly, the advantages of canopy-only optimization for drag performance extend to the capsule-canopy two-body system, enhancing the drag performance of the canopy in the two-body system. This strategic approach reduces the transient calculation time for the two-body system, further improving computational efficiency. The method provides a practical and forward-thinking solution for the design of Mars parachutes.

Pumpkin-Based Cookies Formulated from Optimized Pumpkin Flour Blends: Nutritional and Antidiabetic Potentials

The study looked at the optimization of blends of pumpkin pulp flour, pumpkin seed, and pumpkin starch residue for the development and nutritional evaluation of cookies. 3 optimum blends, runs 12, 15, and 16 were selected out of 16 runs using Optimal Model Design of Response Surface Methodology with responses based on proximate analysis. The amino acid profile, anti-diabetic potential, physical properties, and sensory characteristics of the developed cookies were evaluated. The proximate parameters of the cookies were found to significantly (P≤ 0.05) increased with the levels of pumpkin seed. The carbohydrate content of the pumpkin-based Cookies (47.20-49.12 %) was significantly lower than that of the control, 100 % Wheat Flour (60.29 %). The ash content ranged from 6.20-8.19 %, and dietary fibre, 24.39-26.9 %. The amino acid profile compared well with FAO/WHO recommendations for healthy adults. The most abundant amino acids in the samples were glutamic acid, alanine, leucine, aspartate, and valine. The In-vitro Protein Digestibilities of the pumpkin-based cookies were significantly higher than that of the control sample (WFC). The pumpkin-based cookies effectively inhibited α-amylase and α-glucosidase at 59 % and 69 % inhibition rates respectively. The Glycemic Indices of the pumpkin-based cookies were in the range of 50-52 % compared to the control (WFC) which was 61.43 %, thus registering the pumpkin-based cookies as low-glycemic foods. Sensory evaluation revealed that cookies made from 60 % pumpkin pulp, 24.5 % pumpkin seed, and 14.4 % pumpkin starch residues were generally acceptable as best among the pumpkin-based cookies.

Segmentation of hyphae and yeast in fungi-infected tissue slice images and its application in analyzing anti-fungal blue light therapy.

Candida albicans (C. albicans) is a pathogenic fungus that undergoes morphological transitions between hyphal and yeast, adapting to diverse environmental stimuli and exhibiting distinct virulence. Existing researches on anti-fungal blue light (ABL) therapy have either focused solely on hyphae or neglected to differentiate between morphologies, obscuring potential differential effects. To address this gap, we established a novel dataset of 150 C. albicans-infected mouse skin tissue slice images with meticulously annotated hyphae and yeast. Eleven representative convolutional neural networks were trained and evaluated on this dataset using seven metrics to identify the optimal model for segmenting hyphae and yeast in original high pixel size images. Leveraging the segmentation results, we analyzed the differential impact of blue light on the invasion depth and density of both morphologies within the skin tissue. Comparison results demonstrated that U-Net-BN emerged as superior segmentation accuracy compared to other models, achieving the best overall performance. While both hyphae and yeast exhibited significant reductions in invasion depth and density at the highest ABL dose (180J/cm2), only yeast was significantly inhibited at the lower dose (135J/cm2). This novel finding emphasizes the importance of developing more effective treatment strategies for both morphologies.

Role of Education and Awareness Programs in Fostering Energy Conservation Behavior in Cities: Empowering Urban Sustainability Using Deep Learning Approach

This paper introduces a pioneering framework that emphasizes the critical role of education in promoting energy conservation behavior within urban environments. By synergizing deep learning methodologies with education-awareness formulations and algorithm, the framework aims to bolster energy conservation security in cities. A central aspect of this framework is the development of an advanced Intrusion Detection System (IDS) model, utilizing Restricted Boltzmann Machines (RBMs) tailored to enhance energy consumption security. Through six strategic integrations of educational components, the IDS not only identifies but also mitigates potential threats to energy systems with unprecedented efficiency. Additionally, the research introduces a novel evolutionary feature selection method based on Memetic Algorithms (MAs), which strategically identifies crucial features for RBM training, ensuring optimal model performance. The incorporation of Memetic Algorithms facilitates individual learning and problem-specific knowledge transmission, further enhancing the adaptive capabilities of the system. This multifaceted approach underscores the pivotal role of education in fostering urban sustainability, pushing the boundaries of energy conservation security. By proactively responding to emerging threats, the proposed framework contributes significantly to the overarching goal of sustainable and secure urban energy management.

Ensembles of Multiple Models for Soil Moisture Retrieval from Remote Sensing Data over Agricultural Areas: A Deep Learning-based Framework

In agricultural areas, most surface soil moisture (SM) retrieval models are unstable in terms of their accuracy and performance during crop growth. As a result, there is no consensus on which model performs optimally during the agricultural season. This is because of the uncertainties associated with model physics, initial conditions, model inversion processes, input data, vegetation attenuation and soil characteristics. To better deal with these practical concerns, we propose a simple, but robust SM retrieval method of using combination of multiple models based on deep learning and multi-model ensemble approach, called the DL-MME method, which makes use of the ‘collective intelligence’ and ‘wisdom of crowds’ concepts. The advantages of this method are: (1) robustness to model selection, and (2) robustness to model calibration during the growing season. In addition, this method is less dependent on one type of data across various agricultural areas compared to the single model approach. Firstly, the coupled water cloud model (WCM) and soil backscattering models (Oh model or advanced integral equation model (AIEM)) with different vegetation descriptors were calibrated and validated during the growing season in sugarcane and winter wheat fields for Sentinel-1 backscattering coefficients (VV and VH). SM was also retrieved by employing the trapezoid model (OPTRAM) with different parameters from Sentinel-2 images. To optimize SM retrieval computations, we used the outputs from optical and SAR models, auxiliary features, and reliable in situ SM measurements as inputs to a deep learning convolutional neural network (DL-CNN). For sugarcane and wheat fields in the early stages of crop growth, WCM models retrieved more accurate time-series SM than optical models. OPTRAM soil moisture retrievals showed greater accuracy in the late crop growing season. Time-series SM retrieval accuracy using DL-MME was higher than for the optical and semi-empirical SAR models. According to the results of the in situ validation for wheat (sugarcane) fields, the minimum MAE by an optimal combination of models was around 0.01 (0.02) m3m-3 (RMSE= 0.036 (0.074) m3m-3; R= 0.87 (0.71)). The findings demonstrate that our method is reliable and feasible for SM retrieval. Additionally, our method provides a way to select an optimal model for retrieving time-series SM during the crop growing season.

Railway Accident Causation Prediction with Improved Transformer Model Based on Lexical Information and Contextual Relationships

The railway system is a prime example of a safety-critical system. Predicting the causes of railway accidents holds immense significance in enhancing railway transportation safety. Previous approaches to railway causation analysis have encountered huge challenges regarding data processing and analytical capabilities. To address this concern, this paper proposes an innovative deep model framework based on the Transformer architecture that utilizes historical data on railway equipment accidents to predict the causes behind such incidents. Firstly, this paper proposes the utilization of Convolutional Block Attention in the domain of text processing, serving as a lexical encoder to augment word semantics acquisition in accident texts. Subsequently, in order to address the deficiency of traditional Transformers that lack positional representation information, we propose incorporating a BiGRU (Bidirectional Gated Recurrent Unit) as a contextual positional information encoder to capture contextual positional information in railway accident data effectively. Finally, considering that accident data reports are discrete tabular data, this study suggests employing cue word techniques for preprocessing accident data to alleviate the model's learning burden. We applied the proposed model to the FRA (Federal Railroad Administration) dataset. The results demonstrate that our model surpasses the current state-of-the-art language models, exhibiting superior performance compared to the optimal model with a notable improvement of 3.56%, 0.42%, and 0.76% in Precision, Recall, and F1-score, respectively. Furthermore, our model accurately predicts accident categories prone to misjudgment even when trained on limited data, outperforming existing language models. The study findings will contribute to the prevention and management of railway accidents.