This study investigates the long-term performance degradation and forecasting of three silicon-based photovoltaic technologies—polycrystalline (pc-Si), monocrystalline (mc-Si), and amorphous silicon (a-Si)—using a seven-year dataset (2015–2021) from a semi-arid climate. Degradation rates are quantified through seasonal-trend decomposition and Arrhenius analysis, revealing distinct mechanisms: pc-Si exhibits the lowest annual degradation (0.36%/year), followed by a-Si (0.57%/year), while mc-Si shows the highest (0.77%/year), with a notable thermal annealing effect partially compensating degradation in a-Si. For forecasting performance ratio, four models are compared, where long short-term memory networks achieve the highest accuracy by capturing nonlinear temporal dependencies, while SARIMA offers robust, interpretable results with lower complexity. Beyond predictive performance, the study establishes links between model behavior and underlying physical processes such as degradation and annealing, and analyzes prediction uncertainty in relation to temperature variability and dust accumulation. These findings highlight trade-offs between accuracy, interpretability, and deployment feasibility, providing a framework for PV performance forecasting under univariate, semi-arid conditions, with future work directed toward multivariate, physics-informed approaches across broader technologies and climates.

This study proposes and applies a machine-learning-driven optimization framework to predict and enhance the thermomechanical performance of carbon-free adobe bricks reinforced with straw and sawdust. To move beyond trial-and-error mix design under a strength-insulation tradeoff, the study establishes reproducible mix-selection rules that reduce experimental iterations. Experimental tests show that adding small amounts of straw (1% and 2%) significantly improves compressive strength, increasing it from 5.41 MPa to 9.62 MPa (+78%) and 7.93 MPa (+46.5%), respectively; however, higher dosages lead to a decrease in strength due to excessive porosity. Sawdust reduces mechanical strength but improves insulation by lowering thermal conductivity from 0.632 W/m.K for the reference brick to 0.145 W/m.K at 10% sawdust. Mixed formulations provided the best compromise: with approximately 0.5-4% sawdust and 0.5-4% straw, they maintained compressive strengths above the minimum requirement of 2.07 MPa established by the Mexican adobe construction standard. A measured dataset (density/porosity, Rc/Rf, lambda and Cp) was used to train surrogate models with a 70/15/15 train-validation-test split, 5-fold cross-validation, and grid-search tuning. The machine learning models exhibited distinct predictive capabilities, achieving R2 = 0.323-0.566 for compressive strength and R2 = 0.794-0.991 for thermal conductivity, and multi-objective optimization (Pareto-based selection) further revealed that hybrid mixtures offer the most balanced solutions. These findings confirm the potential of agricultural waste valorization for the production of eco-friendly building materials and establish a systematic methodology that combines experimental work with artificial intelligence to optimize sustainable adobe bricks.

This article introduced a smart energy managementEnergy management strategy that uses two complementary machine learningMachine learning (ML) algorithms to predict and optimize energy consumption patterns in smart environmentsEnvironment. A recurrent neural networkRecurrent Neural Network (RNN) (RNN) is used to model time-seriesTime-series prediction energy data, while an ensemble method (such as XGBoostXGBoost) handles decision-making for energy control actions. A smart energy managementEnergy management strategy is proposed that leverages two complementary machine learningMachine learning algorithms to predict energy consumption and optimize system control. The system is designed to be deployed on low-power embedded devices, enabling local real-time control. Unlike conventional rule-based or static models, our approach incorporates predictive uncertainty and dynamic adaptation, enabling deployment on low-power embedded systemsEmbedded systems or edge devices. The suggested architecture combines a recurrent neural network (RNNRecurrent Neural Network (RNN)) for time series forecasting and a tree-based ensemble method (Random Forest or XGBoostXGBoost) for real-time decision support. The results obtained on publicly available datasets show improved accuracy, faster adaptation, and energy savings compared to baseline single-model approaches. Finally, numerical results demonstrate improved performance over single-model systems in accuracy, robustness, and computational efficiency.

This document has established a new harvesting energy technology for sustainable energy in smart building. However, smart buildings are considered by the combination of several technologies, sensors, and devices designed to capture and analyze data. In addition, building data is used to improve the operational efficiency of buildings, particularly in the management and optimization of resources and services. Additionally, this approach brings together data from a variety of sources, including people, equipment, infrastructure, and systems. Furthermore, these integrated systems contribute to the sustainability of buildings by optimizing energy consumption and minimizing waste. Additionally, the energy harvested can be used to power sensors, appliances, and other infrastructure components, enhancing building intelligence and efficiency. Beyond improving energy efficiency, this approach ensures that energy collectors in public buildings are managed in line with sustainable development objectives, reducing dependence on conventional energy sources. Finally, digital simulation confirms the monitoring and control of essential systems such as energy, lighting, heating, aeration and air conditioning (HVAC). In the same way, the consumption of this public building has been reduced, and its dependence on conventional energy sources is minimized.

This paper introduces an intelligent control strategy based on neural networks to optimize energy consumption in autonomous vehicles. The proposed approach leverages a recurrent neural network (RNNRecurrent Neural Network (RNN)) to forecast power demand in real time, considering vehicle dynamics and environmental conditions, while an ensemble learning model enhances adaptive control decisions. Unlike traditional rule-based energy managementEnergy management systems, this hybrid architecture dynamically adjusts energy usage from multiple sources—including battery systems, regenerative braking, and auxiliary inputs—based on real-time driving contexts such as speed variations, traffic patterns, and terrain changes. Designed for deployment on embedded automotive hardware, the system enables localized, autonomous decision-making with minimal computational overhead. Experimental evaluations on realistic driving datasets demonstrate superior performance in terms of energy efficiency, responsiveness to uncertainty, and overall system stability, validating the potential of neural network-based control for intelligent energy managementEnergy management in next-generation autonomous vehicles.

Jorf, a small town in southeastern region of Morocco, exemplifies durable construction practices derived from ancestral knowledge that should not be overlooked in the era of climate change. To achieve net-zero carbon across the building lifecycle, it is essential to valorize this heritage that addresses both embodied and operational carbon considerations. Within this framework, an insitu experimentation and monitoring campaign was conducted on a single-family rammedearth dwelling in Jorf city. The results indicate that, as the roof albedo increased from 0.55 to 0.85, the maximum indoor air temperature decreased by up to 2.57 degrees C, reaching 36.8 degrees C when the outdoor temperature rose to 50 degrees C. Furthermore, the average daily indoor temperature fluctuation range decreased from 1.7 degrees C to 1.2 degrees C, while the average daily diurnal ambient temperature fluctuation was around 18.7 degrees C during the monitored summer period. Moving forward, three building energy models, i.e., conventional building representing the low-performance scenario, ecological building representing the case study prototype, and ecological building with high albedo coatings, were simulated across 15 cities considering occupied scenario. Interestingly, the results indicate that ecological building reduces peak indoor temperatures by 4.1-7.2 degrees C and showed cooling energy savings of 39.4-68.5 kWh/m2.y, compared to conventional building. Furthermore, high-albedo roofs offer additional benefits, reducing peak indoor temperatures by 1.9-2.3 degrees C and achieving cooling thermal energy savings of 18.47 to 26.13 kWh/m2.y, with respect to ecological scenario. Besides, the combined strategies ensure thermal comfort improvements of 658 h to 1559 h, depending on the studied city. Nonetheless, it is important to consider the countereffects of high-albedo roofs on heating demand, winter discomfort hours, and overall indoor thermal comfort, as the cooling gains achieved in summer can be largely offset during the colder period. Therefore, climate-responsive strategies should aim to balance the risks of overheating and underheating, ensuring optimal thermal comfort throughout the year.

This paper presents a neural network-based control strategy designed to achieve maximum power tracking (MPT) in smart public buildings. Unlike conventional rule-based energy systems, our approach integrates a recurrent neural network (RNNRecurrent Neural Network (RNN)) for real-time forecastingReal-time forecasting of power demand and environmental dynamics, coupled with an ensemble learning model to optimize control actions. The system aims to maximize the extraction and utilization of renewable energy sources (such as solar or piezoelectricPiezoelectric systems), while dynamically responding to changing load profiles and building occupancy patterns. We demonstrate that the proposed control architecture enhances power utilization efficiency, ensures system robustness under uncertainty, and supports autonomous operation on embedded edge devices. Experimental evaluation on realistic datasets confirms improved energy stability, tracking precision, and fault tolerance compared to traditional predictive models.

The increasing demand for cost-effective and efficient renewable energy solutions presents significant optimization challenges in hybrid energy systems. This paper addresses these challenges by conducting a comparative analysis of three advanced optimization algorithms-L & eacute;vy Flight Optimization (LFO), Archimedean Optimization (AO), and Quantum Gorilla Optimization (QGO)-to minimize the Total Net Present Cost (TNPC) and Levelized Cost of Energy (LCOE) in hybrid renewable energy systems. The study integrates critical cost parameters such as Capital Expenditure (CAPEX), Operational Expenditure (OPEX), replacement costs, and salvage values into an advanced optimization framework. Three system configurations are evaluated: Wind Turbines and Fuel Cells (WT/FC), Photovoltaic Systems and Fuel Cells (PV/FC), and a combined system (PV/WT/FC), under varying availability levels (100%, 96%, and 92%). The results demonstrate that LFO consistently outperforms the other algorithms, achieving the lowest TNPC of $0.051 for the WT/FC system at 96% availability, compared to $0.719 using QGO. These findings underscore the importance of selecting tailored optimization strategies to balance cost, performance, and system reliability. This research provides valuable insights into designing efficient and economically viable renewable energy systems, particularly, for applications requiring consistent high energy output, such as monocrystalline and polycrystalline PV-based configurations.

In order to meet sustainability goals in the building sector, this study explores the potential of alfalfa (luzerne) fiber–reinforced adobe bricks, an abundant agricultural by-product, as a sustainable construction solution suitable for Morocco’s warm and semi-arid regions. A comprehensive experimental campaign was conducted to evaluate the physical and thermal properties of these bricks with varying fiber contents ranging from 0

The reliable long-term prediction of photovoltaic (PV) system performance is essential for optimizing operation, maintenance, and energy management in smart cities. Digital Twin (DT) technology has emerged as a promising approach for real-time monitoring and predictive analytics by continuously integrating physical system measurements with virtual models. However, many existing DT-based studies rely on limited validation periods, lack comprehensive uncertainty quantification, and provide insufficient comparisons with conventional forecasting approaches. To address these limitations, this study proposes a data-driven Digital Twin framework for daily photovoltaic energy production prediction for three silicon photovoltaic technologies (amorphous silicon, polycrystalline silicon, and monocrystalline silicon) operating under semi-arid climatic conditions in Morocco. The framework integrates data quality assessment and statistical production trend analysis based on linear regression, bootstrap confidence intervals, the Mann–Kendall trend test, and Sen’s slope estimator. In addition, the proposed DT model was benchmarked against persistence, linear regression, Random Forest, XGBoost, LightGBM, and Long Short-Term Memory (LSTM) models using a chronological training, validation, and independent testing framework. Model performance was evaluated over an independent test period of 602 consecutive days, representing approximately 1.65 years of continuous operation and more than one complete annual cycle. The proposed Digital Twin improved upon the standalone LightGBM model, achieving an RMSE of 1.3047 kWh/day, an MAE of 0.9255 kWh/day, a MAPE of 14.62%, and an R2 of 0.7015, compared with an RMSE of 1.3349 kWh/day, an MAE of 0.9637 kWh/day, and an R2 of 0.6875 for LightGBM. Over the 11-year monitoring period, the estimated long-term production trend rates were −0.566% yr−1 for a-Si, −0.335% yr−1 for pc-Si, and −0.260% yr−1 for mc-Si. Bootstrap analysis yielded median long-term production trend rates of −0.573, −0.326, and −0.251% yr−1, respectively, while Mann–Kendall tests indicated no statistically significant monotonic production trend for any technology (p = 0.1611, 0.2758, and 0.3502, respectively). The extended independent testing period, combined with statistical production trend analysis and comparative machine-learning evaluation, demonstrates the applicability of the proposed Digital Twin framework for photovoltaic performance monitoring and adaptive prediction under semi-arid climatic conditions.

This study integrates a validated three-dimensional non-isothermal computational-fluid-dynamics (CFD) model of a branched-serpentine proton exchange membrane fuel cell (PEMFC) with a calibrated reduced-order model, a neural-network surrogate and multi-objective optimization. The CFD baseline was validated against experimental polarization and power-density data with a maximum deviation below 4%, giving 1.03 A/cm² and 0.515 W/cm² at 0.5 V. A 3000-point Latin hypercube design was then generated from a semi-empirical model recalibrated to this validated baseline and used to train a 128–128–64–32 multilayer perceptron. On the held-out set, R2 values were 0.99994, 0.99997 and 0.99997 for current density, power density and voltage efficiency, respectively; an independently generated 1024-point Sobol design produced comparable accuracy. NSGA-II identified a compromise at Vcell = 0.550 V, Top = 333.15 K, xO2 = 0.205, RH = 1.0, vcathode = 1.40 m/s and vanode = 0.197 m/s, predicting 0.526 W/cm², 44.70% voltage efficiency and a calibrated thermal-gradient proxy of 4.48 °C. Local perturbation and residual-bootstrap analyses showed robust efficiency and thermal-uniformity improvements, whereas the modest power gain was less robust. A common-budget benchmark against multi-objective differential evolution (MODE) and multi-objective particle swarm optimization (MOPSO) further showed that MODE achieved the largest hypervolume. The framework therefore provides a transparent CFD-informed screening layer, while final Pareto candidates still require direct 3D CFD and experimental confirmation.

Employing machine learning (ML) for predictive maintenance represents a promising approach for enhancing the reliability and efficiency of critical water infrastructure, particularly in reverse osmosis (RO) desalination plants where high-pressure pumps, which play a critical role in RO operation, operate under demanding industrial conditions to address growing water scarcity. Against this backdrop, the objective of this study is to develop and validate an industrial ML-based predictive maintenance framework for early fault prediction in high-pressure pumps operating in RO desalination plants. To achieve this, real-time sensor data from 5000-labeled observations including vibration, temperature, pressure, displacement, and electrical parameters were employed in combination with six classification algorithms: random forest (RF), support vector machine, k-nearest neighbors, decision tree, artificial neural network, and naive Bayes. Furthermore, a comparative analysis was conducted to assess the performance of these algorithms, considering accuracy, precision, recall, F1-score, and AUC metrics. Results show that the RF model demonstrated superior performance, achieving 97.3% overall accuracy, a weighted F1-score of 0.97, and an AUC of 0.96, with particularly strong performance on minority fault classes (F1-scores ranging from 0.85 to 0.94). These findings highlight the potential of integrating multisensor industrial monitoring and ML-based analysis for predictive maintenance applications in RO desalination infrastructures operating under real industrial conditions.

This study develops aGreen hydrogen detailed mathematical model for alkaline water electrolyzer cells to simulate and analyze their performance, incorporating factors such as gas evolution, dissolution, bubble formation, and charge transport through a two-phase Euler-Euler method approach. Validated against experimental data at various current densities, the model assesses the effects of potassium hydroxide concentration, separator porosity, and electrolyte flow on flow dynamics and bubble behavior. Gas bubbles formation at electrodes decreases ionic conductivity and available surface area for reactions, increasing overpotential. The study also highlights the use of neural networks and ensemble tree models for predicting hydrogen production rates, achieving an average R-squared value of 0.98, indicating a strong prediction accuracy and potential for process optimization.

Accurate temperature prediction of permanent magnet synchronous motors (PMSMs) is critical for ensuring operational efficiency and reliability. This study investigates the impact of Feature Dimensionality Reduction (FDR) techniques on the performance of Multi-Layer Perceptron (MLP)-based models for PMSM temperature prediction. We compare three approaches: training the MLP model on the original dataset without dimensionality reduction, using Correlation Analysis (CA) to retain only the most informative features, and using Principal Component Analysis (PCA) to project the data onto a reduced feature space. The models are evaluated using widely validated performance metrics, including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Coefficient of Determination (R2). In addition, the execution time for different dataset configurations is analyzed to evaluate the trade-offs between dimensionality reduction and computational efficiency. The results provide a comprehensive analysis of how FDR techniques affect prediction accuracy and processing time and offer insights into optimizing neural network-based predictive models for PMSM health monitoring and diagnosis.

This paper establishes the emergence of AIArtificial Intelligence (AI) technologies tailored to real-world challenges and explores a promising avenue for specific energy harvestingEnergy harvesting methods, notably the optimization of piezoelectricPiezoelectric and triboelectric nanogenerators in terms of recognition, production, and consumption. It proposes an adaptive approach using machine learning and neural network techniques to collect and optimize vibration energy. Recent advancements in computational methodologies, especially within artificial intelligence (AIArtificial Intelligence (AI)) and machine learning (MLMonolayer (ML)), have heightened the need for intelligent, self-sustaining devices. Given the global concern over energy consumption, there is an urgent need for solutions that reduce energy usage while maintaining the efficiency of intelligent applications. Energy harvestingEnergy harvesting technology, which harnesses ambient mechanical vibrations to generate electrical energy, presents a viable solution. This paper aims to demonstrate how cutting-edge AIArtificial Intelligence (AI)-driven approaches can enhance energy harvestingEnergy harvesting efficiency, thereby contributing to the development of sustainable and intelligent energy solutions. Additionally, multiple simulations conducted using MATLAB/Simulink, alongside experimental results obtained with the dSPACE DS1104 board, are discussed to validate the improvements in control and speed estimation.

This paper focuses on the comprehensive evaluation of the mechanical performanceMechanical performances of adobe bricksAdobe bricks fabricated from two distinct clay soils sourced from the Beni Mellal region. The study aims to compare the mechanical strengths of adobes crafted from these clay soils, shedding light on their structural suitabilityStructural suitability for construction practices. The research employs a dual-pronged approach. First, physical testing of adobe bricksAdobe bricks is carried out to ascertain their compression and flexural strengths. The findings reveal that the highest recorded values for compression and flexural strengths are 14.57 and 0.5 MPa, respectively. This empirical data serves as a basis for evaluating the load-bearing capabilities of adobe bricksAdobe bricks, an essential aspect of their feasibility as building components. By investigating these mechanical behaviors, the study provides insights into the structural response and durability of the adobe bricksAdobe bricks, allowing for a comprehensive understanding of their performance under different conditions. This research paper offers a thorough examination of the mechanical properties of adobe bricksAdobe bricks fabricated from clay soils in the Beni Mellal region. By comparing their mechanical strengths and interpreting the factors influencing their performance, the study contributes to an understanding of the suitability of earth materialsEarth materials in construction. With sustainability becoming a key focus in modern construction, the findings of this research offer valuable insights for architects, engineers, and policymakers.

This paper presents a new approach to optimizing the use of electrical energy in smart buildings using neural network techniques. The proposed method has two main objectives: (i) to achieve energy efficiency through the application of intelligent load management; (ii) to ensure stable building system operation through the prediction of energy demand trends. A neural network-based control algorithm was developed and implemented in the building electrical systems for real-time energy optimization. This approach was validated by applying it to a smart building case, with a real-time monitoring and control prototype. Comprehensive analysis, including simulation and experiment tests, validated the system's effectiveness in reducing energy consumption while maintaining operational stability and being an effective solution to sustainable building energy management.

The exploration of bio-based materials for sustainable construction practices, particularly through the use of locally sourced resources like sawdust, is the focus of this study, which evaluates the thermal performance of adobe bricks reinforced with 2% sawdust in small, medium, and large sizes, ranging from 0.3 to 2 cm. The bricks were manufactured using local materials, and the physicochemical properties of the clay were initially analyzed and characterized. The laboratory examined the bricks' thermophysical characteristics, such as their density, thermal conductivity, and heat capacity. The TRNSYS program was used to conduct annual thermal simulations based on representative meteorological data for a typical building in the semi-arid Moroccan city of Beni Mellal. The results indicate that the energy savings achieved in terms of both heating and cooling were comparable across all configurations of sawdust-reinforced adobe bricks. When compared to a reference concrete building, the heating energy demand was reduced by 59.14% for clay without sawdust, 72.61% for clay with small sawdust, 71.25% for clay with medium sawdust, and 69.88% for clay with large sawdust. Similarly, the cooling energy demand reductions were 45.71%, 58.82%, 57.68%, and 56.62%, respectively, for clay without sawdust, and with small, medium, and large sawdust. These findings suggest that the incorporation of sawdust, regardless of particle size, leads to similar energy savings, offering flexibility in utilizing locally available sawdust from Beni Mellal to optimize energy performance. This research highlights the importance of local clay-based materials and the use of sawdust as a natural reinforcement, to optimize the energy efficiency of buildings. It also offers new perspectives for their integration into sustainable construction practices, contributing to global sustainable development goals.

As awareness of the environmental impacts of traditional building materials like cement and concrete—known for their significant greenhouse gas emissions and high energy requirements for heating and cooling—increases, there is growing interest in adopting eco-friendly and sustainable alternatives. This study evaluates the thermal performance of clay brickClay brick constructions across various climatic zones in Morocco, emphasizing their impact on indoor thermal comfortThermal comfort and energy efficiency. Using OpenStudio software, we analyzed annual indoor temperature variations and energy consumption for heating and cooling in six distinct cities: Rabat, Tangier, Fes, Ifrane, Benguerir, and Ouarzazate. Our results demonstrate that clay brickClay brick constructions significantly enhance thermal comfortThermal comfort. In Rabat, up to 46

The environmental problemsOffice wastepaper causedClay brick byComposite material the use of conventional building materials, such as cement, and their rapid depletion, make it necessary to develop environmentally friendly building materials made from renewable resources. Inadequate management of waste office paper in landfill sites leads to air and groundwater pollutionPollution. In this context, the exploitation of wastepaper as a renewable resource, with its use as a reinforcement in the manufacture of unfired adobe bricksAdobe bricks, appears to be a solution that can resolve these environmental problems in a more environmentally friendly way. In this study, waste office paper was used as an additive in adobe clay bricksClay brick, with weight proportions of 0, 2, 4, and 10