Traditional Simulation Research Articles

Simulation method for filling stage in liquid composite molding based on a two-phase flow volume of fluid model

Liquid composite molding (LCM) is a process for manufacturing fiber-reinforced resin matrix composites. The chemical rheological properties of resin in the filling stage of LCM will produce a non-isothermal flow process that includes multi-field coupling of thermal field, flow field, and chemical field. During the filling stage, the resin is injected into a closed mold pre-laid with a fiber at a certain temperature and pressure to flow continuously and undergo a curing chemical reaction. The viscosity of resin changes constantly under the influence of temperature and the curing reaction, which affects the flow of resin. The simulation of this multi-field coupling filling stage in LCM can provide an effective reference for the design of process parameters. The simulation method of the filling stage in LCM is studied in this paper. The mathematical model of the filling stage is derived based on the air–resin two-phase flow volume of the fluid model. Based on the secondary development of the ANSYS FLUENT software, the mathematical model is solved to realize the simulation. The parameters of an alkali-free glass fiber plain weave fabric are measured and the calculation model of thermosetting resin viscosity is established by designed experiments, which are applied in simulation. The filling stage experiment of LCM is designed to verify the accuracy of simulation results. Meanwhile, by comparing with the traditional simulation method, it is found that the data error obtained by the simulation method in this paper is significantly reduced.

Enhancing low-temperature CO removal in complex flue gases: A study on La and Cu doped Co3O4 catalysts under real-world combustion environment

Designing CO oxidation catalysts for complex flue gases conditions is particularly challenging in fire scenarios. Traditional flue gas simulations use a few representative gases but often fail to adequately evaluate catalyst performance in real-world combustion conditions. In this study, we developed doping strategies using La and Cu to enhance the water resistance of Co3O4 catalysts. Catalyst 0.1La-Co3O4-CuO/CeO2 exhibits exceptional low-temperature catalytic activity, achieving 100% conversion at 130 °C. This enhancement is largely due to the introduction of La, which increases the active Co3+/Co2+ ratio and suppresses hydroxyl group formation on the Co3O4 surface. Cu doping also changes the Co3O4 lattice structure, forming Cu+ as active sites and enhancing the activity at low temperatures. For the first time, steady-state tube furnace and fixed bed were employed to evaluate the catalytic performance of CO in actual combustion atmosphere. Catalyst 0.1La-Co3O4-CuO/CeO2 maintains excellent catalytic efficiency (T100 = 120 °C) under well-ventilated conditions. However, its activity significantly decreases in poorly ventilated environments, due to the competitive adsorption of small molecules at active sites, such as acetone, commonly found in smoke. This study provides valuable insights for designing water-resistant, low-temperature, non-noble metal catalysts and offers a methodology for evaluating CO catalytic activity in real-world environments.

Efficient two-stage modeling of heat plume interactions of geothermal heat pumps in shallow aquifers using convolutional neural networks

This paper presents an innovative approach to model the impact of geothermal heat pumps on groundwater temperature and the interaction of multiple heat plumes within the aquifer. The significance of this research lies in its applicability to real-time urban planning through a web application, where traditional simulation methods prove too time-consuming. Our methodology involves a two-stage neural network approach, leveraging convolutional neural networks (CNNs), trained on a simulated dataset constructed with realistic subsurface flow parameters extracted from borehole measurements in the Munich region.The research process is systematically structured into stages: the first stage rapidly predicts the general shape of a single heat pump’s plume in the absence of other influences. Building upon this, the second stage refines predictions by incorporating the interactions with neighboring heat plumes. Both stages employ CNNs, allowing for efficient training and evaluation. Experiments are conducted on dataset sizes of 100 and 1000 data points and on input parameters, focusing on a reduced dataset with data points of a spatial size of 256x16 pixels. The dataset is split into 70% training, 20% validation and 10% test data.Our results showcase the effectiveness of our approach, achieving a Root Mean Square Error (RMSE) of approximately 0.1 °C in both stages. This level of accuracy demonstrates the viability of our two-stage model in capturing the complex dynamics of GHWP interactions in shallow aquifers. The proposed methodology not only outperforms known analytical approximations but also significantly reduces computational costs of simulations, making it a promising tool for practical applications in city planning and beyond.

Predicting the Structure of Enzymes with Metal Cofactors: The Example of [FeFe] Hydrogenases.

The advent of deep learning algorithms for protein folding opened a new era in the ability of predicting and optimizing the function of proteins once the sequence is known. The task is more intricate when cofactors like metal ions or small ligands are essential to functioning. In this case, the combined use of traditional simulation methods based on interatomic force fields and deep learning predictions is mandatory. We use the example of [FeFe] hydrogenases, enzymes of unicellular algae promising for biotechnology applications to illustrate this situation. [FeFe] hydrogenase is an iron-sulfur protein that catalyzes the chemical reduction of protons dissolved in liquid water into molecular hydrogen as a gas. Hydrogen production efficiency and cell sensitivity to dioxygen are important parameters to optimize the industrial applications of biological hydrogen production. Both parameters are related to the organization of iron-sulfur clusters within protein domains. In this work, we propose possible three-dimensional structures of Chlorella vulgaris 211/11P [FeFe] hydrogenase, the sequence of which was extracted from the recently published genome of the given strain. Initial structural models are built using: (i) the deep learning algorithm AlphaFold; (ii) the homology modeling server SwissModel; (iii) a manual construction based on the best known bacterial crystal structure. Missing iron-sulfur clusters are included and microsecond-long molecular dynamics of initial structures embedded into the water solution environment were performed. Multiple-walkers metadynamics was also used to enhance the sampling of structures encompassing both functional and non-functional organizations of iron-sulfur clusters. The resulting structural model provided by deep learning is consistent with functional [FeFe] hydrogenase characterized by peculiar interactions between cofactors and the protein matrix.

Advancing Traffic Simulation Precision and Scalability: A Data-Driven Approach Utilizing Deep Neural Networks

In traditional traffic simulation studies, vehicle behavior has typically been modeled using complex analytical frameworks, which often struggle to encompass the full range of variables affecting vehicle operations. Addressing this gap, our research introduces an innovative data-driven framework for traffic simulation that incorporates human driving data into its decision-making processes. This enables the modeling of diverse vehicle behaviors by taking into account both vehicle-specific characteristics and environmental factors. At the core of this framework are two advanced deep neural networks, convolutional long short-term memory and convolutional gated recurrent unit, which underpin our vehicle traffic simulation model. Utilizing datasets from the Next Generation Simulation project, specifically the I-80 and US-101 road sections, our study further evaluates the framework’s performance through single-step continuous prediction, as well as transferability tests, employing the TransMSEloss function to optimize prediction accuracy. Our findings reveal that the proposed data-driven model significantly outperforms traditional models, achieving an exceptional accuracy of 97.22% in training and 95.76% in testing. Notably, in continuous prediction, our model maintains an 89.57% accuracy up to the fifth step, exceeding the traditional framework’s 82.82% by 5% to 10% at each step. Time cost analysis indicates that while the data-driven framework’s advantages are more pronounced in large-scale simulations, it also demonstrates strong transferability, with a 93.48% accuracy on diverse datasets, showcasing its applicability across different traffic scenarios. This study not only highlights the potential of deep learning in traffic simulation, but also sets a new benchmark for accuracy and scalability in the field.

An aero-structure-acoustics evaluation framework of wind turbine blade cross-section based on Gradient Boosting regression tree

Under the rapid development of wind turbines, the rotor size has substantially increased in recent years. To meet the key design criteria, finding a trade-off between aerodynamic, structure and noise (ASN) impact becomes a challenging problem. The blade cross-section is the basic element of the blade, its outer contour is the airfoil profile that produces aerodynamic loads as well as noise, and the inner part is the supporting composite material that provides enough stiffness to balance the loads. Modifying local blade sections can adjust the rotors’ overall performance. However, in the work of blade-combined ASN optimization, the iterative process using traditional numerical simulation methods becomes extremely heavy. In this study, a sustainable database is created based on a large number of calculations of aerodynamic, structural and noise attributes at various cross-sections. Then a platform named AFML (Airfoil machine learning) for simultaneously predicting the comprehensive performance of the cross-section is constructed by using the integrated Gradient Boosting Regression Tree algorithm. Results show that the prediction accuracy of AFML is acceptable even for unseen inflow conditions. By calling the pre-trained AFML, ASN data of the cross-section can be immediately obtained, and the blade shape and inner structure can be updated quickly.

Rapid Cycle Deliberate Practice vs Traditional Simulation for Neonatal Resuscitation Training and Retention of Skills

Rapid Cycle Deliberate Practice vs Traditional Simulation for Neonatal Resuscitation Training and Retention of Skills

A Temperature Prediction Model for Flexible Electronic Devices Based on GA-BP Neural Network and Experimental Verification.

The problem that the thermal safety of flexible electronic devices is difficult to evaluate in real time is addressed in this study by establishing a BP neural network (GA-BPNN) temperature prediction model based on genetic algorithm optimisation. The model uses a BP neural network to fit the functional relationship between the input condition and the steady-state temperature of the equipment and uses a genetic algorithm to optimise the parameter initialisation problem of the BP neural network. To overcome the challenge of the high cost of obtaining experimental data, finite element analysis software is used to simulate the temperature results of the equipment under different working conditions. The prediction variance of the GA-BPNN model does not exceed 0.57 °C and has good robustness, as the model is trained according to the simulation data. The study conducted thermal validation experiments on the temperature prediction model for this flexible electronic device. The device reached steady state after 1200 s of operation at rated power. The error between the predicted and experimental results was less than 0.9 °C, verifying the validity of the model's predictions. Compared with traditional thermal simulation and experimental methods, this model can quickly predict the temperature with a certain accuracy and has outstanding advantages in computational efficiency and integrated application of hardware and software.

In silico agent-based modeling approach to characterize multiple in vitro tuberculosis infection models.

In vitro models of Mycobacterium tuberculosis (Mtb) infection are a valuable tool for examining host-pathogen interactions and screening drugs. With the development of more complex in vitro models, there is a need for tools to help analyze and integrate data from these models. To this end, we introduce an agent-based model (ABM) representation of the interactions between immune cells and bacteria in an in vitro setting. This in silico model was used to simulate both traditional and spheroid cell culture models by changing the movement rules and initial spatial layout of the cells in accordance with the respective in vitro models. The traditional and spheroid simulations were calibrated to published experimental data in a paired manner, by using the same parameters in both simulations. Within the calibrated simulations, heterogeneous outputs are seen for bacterial count and T cell infiltration into the macrophage core of the spheroid. The simulations also predict that equivalent numbers of activated macrophages do not necessarily result in similar bacterial reductions; that host immune responses can control bacterial growth in both spheroid structure dependent and independent manners; that STAT1 activation is the limiting step in macrophage activation in spheroids; and that drug screening and macrophage activation studies could have different outcomes depending on the in vitro culture used. Future model iterations will be guided by the limitations of the current model, specifically which parts of the output space were harder to reach. This ABM can be used to represent more in vitro Mtb infection models due to its flexible structure, thereby accelerating in vitro discoveries.

Physics‐informed learning of chemical reactor systems using decoupling–coupling training framework

AbstractIt is known that physics‐informed learning become a new learning philosophy that has been applied in many scientific domains. However, this approach often struggles to achieve optimal performance in addressing the issue of multiphysics coupling. Here, for the first time, we extend this approach to modeling chemical reactor systems. We design a new decoupling–coupling training framework, which consists of decoupling pre‐training and multiphysics coupling training steps. With decoupling pre‐training, the complex physical domain is decomposed into subdomains of fluid flow, heat transfer, and mass transfer combined with reaction kinetics. Each subdomain is represented by a specialized neural network that can provide a coarse but reasonable distribution of network parameters for initializing the sub‐networks for the subsequent multiphysics coupling training. The capabilities of this approach, in comparison with the traditional CFD simulation, are demonstrated through an example of a plate reactor system with a heating cylinder.

Consistent Construction of the Density Matrix from Surface Hopping Trajectories.

Proper construction of the density matrix based on surface hopping trajectories remains a difficult problem. Due to the well-known overcoherence in traditional surface hopping simulations, the electronic wave function cannot be used directly. In this work, we propose a consistent density matrix construction method, which takes the advantage of occupation of active states to rescale the coherence calculated by wave functions and ensures the intrinsic consistency of the density matrix. This new trajectory analysis method can be used for both Tully's fewest switches surface hopping (FSSH) and our recently proposed branching corrected surface hopping (BCSH). As benchmarked in both one- and two-dimensional standard scattering models, the new approach combined with BCSH trajectories achieves highly accurate time-dependent spatial distributions of adiabatic populations and coherence compared to exact quantum results.

Advanced VLSI-based digital image contrast enhancement: A novel approach with modified image pixel evaluation logic

This research uses an extremely memory-intensive and power-efficient very-large-scale-integration (VLSI) architecture to perform nonlinear scaling on pictures. Images can be scaled either up or down so that they can be adapted to the differing resolutions of a variety of devices, such as cameras and printers. Photography with a high resolution calls for significantly more storage space and processing power. The branch of image processing known as “image fusion” has recently had a meteoric rise in prominence. The rapidly growing usage of digital imagery in remote sensing and satellite applications has led to an increase in the need to both store and process massive amounts of picture data. This demand has resulted in an increase in the availability of storage space. Traditional fusion methods produce spatial distortions despite their generally high level of spatial performance. Because of this, the discrete wavelet transformation (DWT), currently the most efficient approach for processing images, focuses on finding a solution to this problem. Within the scope of this investigation, we propose the modified image pixel evaluation logic (MIPEL), a functional method for incorporating the lifting-based strategy utilized in DWT-based picture fusion. In order to avoid excessive delays, high-resolution photo processing typically makes use of hardware processing, which necessitates the utilization of more complex apparatus and more time. In this work, the hardware description language verilog is employed to evaluate the effectiveness of various fundamental approaches to the enhancement of pictures. VHDL is a relatively recent method that has replaced more traditional simulations in the process of producing results for signal processing. It does this by allowing rapid access to physical VLSI representations. Utilizing the proposed MIPEL as a starting point, this study aims to design, model, simulate, and develop several different strategies for enhancing photos using FPGA. Because they take place on the pixel level and in the immediate neighborhood of a pixel, the enhancements that may be made to an image can be thought of as point processing procedures. VLSI architecture is developed for contrast augmentation for low-contrast source photos using pixel-based image fusion, and in the simulation results that followed, we offer findings that clearly illustrate the efficiency of this technology.

Cosmic variance suppression in radiation-hydrodynamic modelling of the reionization-era 21-cm signal

ABSTRACT The 21-cm line emitted by neutral hydrogen is the most promising probe of the Epoch of Reionization (EoR). Multiple radio interferometric instruments are on the cusp of detecting its power spectrum. It is therefore essential to deliver robust theoretical predictions, enabling sound inference of the coeval Universe properties. The nature of this signal traditionally required the modelling of $\mathcal {O}(10^{7-8} \, {\rm Mpc}^3)$ volumes to suppress the impact of cosmic variance. However, the recently proposed Fixed & Paired (F&P) approach uses carefully crafted simulation pairs to achieve equal results in smaller volumes. In this work, we thoroughly test the applicability of and improvement granted by this technique to different observables of the 21-cm signal from the EoR. We employ radiation-magneto-hydrodynamics simulations to ensure the most realistic physical description of this epoch, greatly improving over previous studies using a seminumerical approach without accurate galaxy formation physics and radiative transfer. We estimate the statistical improvement granted by the F&P technique on predictions of the skewness, power spectrum, bispectrum, and ionized regions size distribution of the 21-cm signal at redshift 7 ≤ z ≤ 10 (corresponding to ${\ge}80{{\ \rm per\ cent}}$ of the gas being neutral). We find that the effective volume of F&P simulations is at least 3.5 times larger than traditional simulations. This directly translates into an equal improvement in the computational cost (in terms of time and memory). Finally, we confirm that a combination of different observables like skewness, power spectrum, and bispectrum across different redshifts can be utilized to maximize the improvement.

Quick large-scale spatiotemporal flood inundation computation using integrated Encoder-Decoder LSTM with time distributed spatial output models

Accurate spatiotemporal flood simulations are essential for making informed decisions regarding flood release in affected regions, such as flood detention areas. Traditional spatiotemporal flood simulation approach uses partial differential equation (PDE) models (or physics-based models), which need high computational time. Although many machine learning (ML) models for inundation are increasingly being used to emulate the PDE models to address this issue, utilizing conventional ML models to achieve large-scale spatiotemporal flood prediction (i.e., simulation output in tens of thousands of grids and time steps over the whole flood event) remains a significant challenge. Therefore, we developed a new inundation model (IM) using encoder-decoder long short-term memory (ED-LSTM) with Time Distributed Spatial output model (ED-LSTM-TDS) that can acquire accurate spatially distributed flood information more rapidly. In the new IM framework, each ED-LSTM-TDS is built to simultaneously generate output for multiple (K) cell grids and multiple ED-LSTM-TDS models are built for prediction at all grids of entire PDE model. This study is the first of its kind to employ the ED-LSTM-TDS method to address spatiotemporal flood inundation simulation problems for flood detention areas. A 1994 km2 flood detention area in northeastern China was used as a case study. ED-LSTM-TDS exhibited better performance in predicting flood characteristics (e.g., water depth, velocity) than alternative methods, including ordinary LSTM, artificial neural network (ANN), and multiple linear regression (MLR). In addition, we investigated the trade-off relationship between the accuracy of flood characteristic prediction and the computation time of the proposed model by considering different numbers of K cell grids in each ED-LSTM-TDS model. The final proposed inundation model could accurately predict the spatiotemporal flood characteristics within 1.5 min to acquire the same information that required approximately 1 h by the PDE models. Such rapid and accurate prediction by the proposed model is important for evacuation planning, upstream flood control decisions and flood damage reduction.

A hybrid MLP-CNN model based on positional encoding for daytime radiative cooler

Radiative cooling is a groundbreaking technology for sustainable temperature regulation to combat global warming and the urban heat island effect. Traditional optimization algorithms for cooling system structures are often brute-force and even limited to local optimal solutions. Here, we propose a joint simulation method based on a multilayer perceptron (MLP) and a convolutional neural network (CNN) with position encoding (PE), namely the hybrid MLP-CNN model with PE. This method can significantly reduce simulation time, avoid getting trapped in local optima, and accurately predict the optical response and radiative cooling power of the structure. Applying PE enables the MLP model to learn the relationship between structural parameters and optical response more quickly and effectively, reducing loss during training and more minor root mean square error (RMSE) and mean relative error (MRE) during testing. In addition, the residual network can reduce the problem of overfitting in the network. Compared to using only an MLP, the hybrid MLP-CNN with PE model can effectively reduce loss and achieve more accurate prediction outputs. The hybrid model is much faster than traditional electromagnetic simulations, taking only 1% of the time, and the model dramatically reduces the time needed to predict the optical response of structures and effectively decreases the cost of designing optical devices. Utilizing this hybrid model not only avoids the potential drawbacks of being limited to suboptimal solutions and significantly reduces the simulation time, but also effectively predicts the optical response of the device. This approach can demonstrate remarkable adaptability in optimizing geometric parameters of various optical apparatuses, including polarization converters, beam splitters, metalens, and other two-dimensional nanostructures.

Field Programmable Gate Array-Based Acceleration Algorithm Design for Dynamic Star Map Parallel Computing

The dynamic star simulator is a commonly used ground-test calibration device for star sensors. For the problems of slow calculation speed, low integration, and high power consumption in the traditional star chart simulation method, this paper designs a FPGA-based star chart display algorithm for a dynamic star simulator. The design adopts the USB 2.0 protocol to obtain the attitude data, uses the SDRAM to cache the attitude data and video stream, extracts the effective navigation star points by searching the starry sky equidistant right ascension and declination partitions, and realizes the pipelined displaying of the star map by using the parallel computing capability of the FPGA. Test results show that under the conditions of chart field of view of Φ20° and simulated magnitude of 2.0∼6.0 Mv, the longest time for calculating a chart is 72 μs under the clock of 148.5 MHz, which effectively improves the chart display speed of the dynamic star simulator. The FPGA-based star map display algorithm gets rid of the dependence of the existing algorithm on the computer, reduces the volume and power consumption of the dynamic star simulator, and realizes the miniaturization and portable demand of the dynamic star simulator.

Confined aquifer dewatering optimization with a modified simulation–optimization method capable of determining the optimal well screen length and depth

Simulation-optimization methods are widely used in dewatering optimization. However, traditional simulation–optimization methods do not address the optimization of well screen length and depth. This study proposes a modified simulation–optimization method for confined aquifer dewatering optimization, which is capable of determining the optimal screen length and depth. The proposed method is based on the linear programming method, and the multivariate adaptive regression splines method is also introduced to develop the prediction model for the parameters required in the linear programming model. A hypothetical case of deep excavation dewatering was optimized using the proposed method to demonstrate its feasibility, with the optimal pumping rate, screen length and depth of each well computed. Furthermore, parametric studies were performed to investigate the effects of some key factors on the optimization results, such as the number of considered pumping wells, required drawdown, insertion ratio of the waterproof curtain, aquifer anisotropy coefficient, and prescribed well screen. The results show that optimal total pumping rate and screen length generally increase with increasing required drawdown and aquifer anisotropy coefficient, while they decrease with increasing well number and insertion ratio of the waterproof curtain. Adjusting screen length is more critical to the optimization results since lower screen depth is always preferred. Optimizing well screen is more essential for higher well number, insertion ratio of the waterproof curtain, and lower aquifer anisotropy coefficient.

Non-volatile double-tunable vortex metalens design based on Sb2S3 using deep neural network and particle swarm optimization algorithm

A dual-tunable vortex metalens is proposed that leverages deep neural network (DNN) and particle swarm optimization algorithm (PSO), incorporating the use of the phase-change material Sb2S3, which enables simultaneous tunable of both the focal length of the focused vortex beam and the topological charge. The control of traditional vortex metalens relies on changing the polarisation state of the incident light; however, its control range is limited. We employed neural network algorithms to learn the complex mapping relationship between the unit structure and the phase, achieving a fast and accurate prediction of the vortex metalens's phase. Compared with traditional simulation methods, the application of this neural network significantly reduces the time consumption of the design process and avoids the need for re-simulation. Furthermore, our design transcends limitations associated with specific wavelengths and enables the rapid design of tunable metalens across the 1–1.5 μm wavelength range, offering flexibility for the development of dual-tunable metalens. Based on this scheme, we simulated vortex metalens with tunable focal length and topological charge, and the results showed that the simulation results were consistent with the theoretical predictions.

A new born‐approximation approach to compute the effects of motion on the solution of electromagnetic problems involving moving materials with stationary boundaries

AbstractThe non‐relativistic motion of media induces very weak effects on electromagnetic fields. For this reason, it is extremely difficult to compute such effects with traditional techniques. To overcome this difficulty, a novel approach is introduced in this work. It is based on the Born approximation of the solution of electromagnetic problems involving media in motion. The formulation of these problems makes use of equivalent field sources and is specifically tailored to find the effects of motion on the electromagnetic field. The new methodology exploits traditional simulators able to deal with media at rest. The motion of the materials has to be managed by specific programs performing simple algebraic calculations. The simplest of the methods proposed does not present any additional complexity with respect to more traditional approaches to the same problems. All methods deal with time‐harmonic electromagnetic fields and, for this reason, they can manage materials in motion with stationary boundaries. A complete set of simulations for cylinders moving in the axial direction show that the new methods outperform traditional numerical approaches and, in some significant cases, traditional approaches based on semi‐analytical techniques. The good features of the new methodology are shown to hold true for a range of velocity values spanning 11 decades; such a range covers most of the applications of practical interest.

Time-dependent photovoltaic performance assessment on a global scale using artificial neural networks

The integration of Renewable Energy Sources (RESs), particularly solar PhotoVoltaics (PVs) has become an imperative aspect of sustainable energy systems. In this pursuit, accurate and efficient simulation tools play a pivotal role in optimizing the performance of PV systems. Traditional simulation approaches, while effective, are often characterized by computational complexities and time-intensive processes. This paper introduces a groundbreaking paradigm in solar energy modeling by harnessing the power of Artificial Neural Networks (ANNs) to revolutionize the accuracy and reliability of PV system simulations. In this work, an hourly, daily, monthly and yearly comparison of the electrical energy obtained with the 5-parameter model and those obtained with the ANNs was developed. For this purpose, a very wide ensemble of localities around the world and types of PV systems were considered in the training and validation phase. ANNs exhibited a maximum mean absolute relative error of 3.5% during training and consistently maintained hourly relative errors below 5% across diverse localities during validation. Hourly power forecasting remains acceptable also in localities with extreme weather conditions. Monthly errors peak at high negative and positive latitudes in summer months when daylight duration exceeds nighttime. However, in the least accurate locality, yearly energy forecasting yielded a maximum error of 8%. Empirical equations based on the trained ANNs are proposed and a relative input-output importance criterion was applied to detect the impact of air temperature and solar radiation on the performance of each PV module. The proposed ANNs demonstrate significant utility in decision-making and real-time processes, providing a valuable framework for managing energy flows within a network and predicting energy production during specific time intervals. This alternative approach surpasses conventional dynamic simulation methodologies found in existing literature in terms of computational cost with comparable accuracy.