Sustainable Energy and Artificial Intelligence

Abstract The proposed solar-assisted hybrid system integrates decentralized, energy-efficient technologies for multi-vector energy production, delivering key advantages such as improved efficiency, lower emissions, economic feasibility, sustainability, and enhanced reliability. This work investigates the optimal design of a solar–fossil-fuel-based configuration capable of generating cooling, heating, power, and freshwater (CCHPW). The system employs different prime movers for combined heat and power production, namely a gas engine (GE), gas turbine (GT), and solid oxide fuel cell (SOFC). The overall plant consists of a prime mover, two types of chillers, an auxiliary boiler, a reverse osmosis desalination unit, parabolic trough solar collectors, a proton exchange membrane (PEM) electrolyzer, and thermal and cooling energy storage systems. A genetic algorithm is utilized to minimize the total annual cost (TAC). Optimization results demonstrate that the exergy efficiency of the GE-based CCHPW system is about 40.31% and 92.50% higher than that of SOFC- and GT-based systems, respectively. Moreover, the GE configuration achieves reductions in TAC of 27.08% and 23.80% compared to SOFC and GT systems, respectively.

Abstract In recent years, Digital Image Correlation (DIC) method has gained attention as a non-contact technique for strain measurement. This study explores the potential of DIC to measure stress-strain behavior of a steel tensile test specimen and calculate the elastic modulus, comparing the results with analytical anticipations and resistive strain gauges measurements. For this purpose, a steel specimen with specified dimensions and material is prepared using white paint spray and subjected to tensile loading. The average absolute error of about 3.7% in strain measurements using resistive strain gauges compared to the analytical results demonstrates the capability of this method to accurately predict strain values. Although the DIC method can predict the linear stress-strain behavior, it shows a 45% error in calculating the elastic modulus compared to both resistive strain gauge measurements and analytical method predictions. To examine the effects of using titanium dioxide powder, preparing the specimen with this powder is recommended to enhance the brightness of the speckles and create sharper edges between the background and the painted speckles. This approach yields an elastic modulus prediction of 163 GPa compared to 183 GPa from the resistive strain gauge method (about a 10% difference), which is acceptable from an engineering perspective. Furthermore, increasing the speckle dimension reduces the prediction error of the elastic modulus to less than 5% compared to the both approaches. The results highlight the high potential of the DIC technique which could be so advantageous in predictive maintenance of energy infrastructure, or artificial intelligence (AI)/ data driven applications.

Abstract Accurate estimation of cellular network throughput is essential for effective network management and user experience optimization. This study conducts a comprehensive comparative analysis of four prominent machine learning (ML) models (i.e., Random Forest Regressor (RFR), Gaussian Process Regressor (GPR), K-Nearest Neighbors (KNN), and Support Vector Regressor) for predicting downlink and uplink data rates based exclusively on the user-base station distance. Evaluation metrics, including coefficient of determination (R2 score), Mean Squared Error (MSE), Mean Absolute Error (MAE), and computational runtime, were employed to assess model performance. Results indicate that ensemble-based (RFR) and probabilistic kernel-based (GPR) approaches outperform instance-based (KNN) and margin-based (SVR) methods in terms of predictive accuracy and error minimization. Furthermore, RFR achieves a favorable balance between accuracy and computational efficiency, making it a practical choice for real-time throughput prediction. Beyond performance estimation, the proposed approach offers potential benefits for energy optimization, enabling more efficient resource allocation both at the user device and base station (or network infrastructure) level. The findings suggest that incorporating additional network parameters and signal quality features could further improve model effectiveness in capturing the complex dynamics of cellular communications. This study employs a public real-world 4G LTE dataset comprising 135 traces (≈15 minutes each) collected under a static mobility model. The proposed distance-only prediction framework highlights that accurate throughput estimation can be achieved using a single distance feature, showing that lightweight models can remain competitive for real-time deployment.

Abstract This study presents a data-driven approach for modeling subgrid-scale (SGS) stresses in turbulent channel flow using a fully connected neural network (FCNN), also known as a multilayer perceptron (MLP). Our data-driven closure model is based on localized learning, and the FCNN architecture has a point-to-point mapping framework. The MLP-based SGS model is trained using direct numerical simulation (DNS) data of a turbulent channel flow at the bulk Reynolds number Re=4400, corresponding to the friction Reynolds number 〖 Re〗_τ=180. Performance of the FCCN model is assessed using a priori analysis in comparison with the filtered DNS (fDNS) data of the turbulent channel flow. Resolved flow statistics, including filtered velocity gradients and wall distance, are employed as input parameters for training the FCNN. To evaluate the significance of the feature inputs, we utilize random forest regression, which reveals that y^+ is a critical factor in predicting the output, which can reduce the compactional costs. In an a priori test, the model achieved a correlation coefficient exceeding 95% for the components of the SGS stress tensor, which have non-zero mean values. Finally, potential strategies for improving prediction accuracy and overall performance are explored. The findings demonstrate that the proposed FCNN can accurately reproduce key turbulence characteristics, offering a promising step toward efficient, AI-based turbulence modeling.

Abstract DC micro-grids (DC-MGs) have attracted significant attention as a promising solution for integrating distributed renewable sources to main utility. Compared to AC systems, DC-MGs offer higher efficiency; however, their varied structures and control strategies introduce considerable dynamic complexity. This paper investigates the dynamic stability of an islanded DC-MG comprising a photovoltaic (PV) unit, battery unit, and the supercapacitor (SC) unit connected to the main bus through DC-DC power converters. The system is modeled using state-space averaging method, and the linearized equations are analyzed using system eigenvalue analysis method to examine the system stability. The impact of various parameters including capacitor size, inductor values, line resistances, and constant power loads (CPLs) on the system's dynamic behavior is verified systematically. Results indicate that larger capacitors and inductors shift eigenvalues further into the left half-plane, improving damping and stability, whereas resistances and CPLs have a destabilizing effect. All simulations and verifications are conducted in MATLAB/Simulink to ensure analytical accuracy. The presented framework offers valuable insights for designing reliable DC-MGs with hybrid energy storage systems for future modern power grids.

Abstract The small and medium scale pumped storage hydro power plants are very useful manner of power generation for safe supply of extra electrical power demand of some urban and rural areas during peak hours in many under development countries. But the electric power consumption of the main transfer pumps of these power plants, which usually work continuously about 12~16 hours on a daily basis, is a considerable amount of electric power that is usually supplied via the national cross country electrical power grid network.
Usually, renewable energy sources utilization are more economic in “Hybrid” combination with other conventional electric power sources in under development countries. Otherwise, they have not significant chances to be utilized and effectively developed in these countries.
In this paper, a typical 1040 MW nominal capacity pumped storage hydro power plant, with a total main pumps power consumption of about 50 MW is studied for seeking opportunity for supply part of their electrical power demand via hybrid power generation of a series of solar PV panels (i.e., a small solar PV power plant) and simultaneous electrical power get from national cross-country power grid network.
This study resulted that about 12% of the main pumps power demand can be generated via a suitable design and proper arrangement of series of solar PV panel arrays