Publications Scientifiques

Permanent URI for this communityhttps://dspace.univ-boumerdes.dz/handle/123456789/10

Browse

Filters

Settings

Subject: search.filters.subject.Artificial neural network

Search Results

Now showing 1 - 10 of 19
  • Thumbnail Image
    Item
    AI-Driven Optimization of Drilling Performance Through Torque Management Using Machine Learning and Differential Evolution
    (Multidisciplinary Digital Publishing Institute (MDPI), 2025) Boukredera, Farouk Said; Hadjadj, Ahmed; Youcefi, Mohamed Riad; Ouadi, Habib
    The rate of penetration (ROP) is the key parameter to enhance drilling processes as it is inversely proportional to the overall cost of drilling operations. Maximizing the ROP without any limitation can induce drilling dysfunctions such as downhole vibrations. These vibrations are the main reason for bottom hole assembly (BHA) tool failure or excessive wear. This paper aims to maximize the ROP while managing the torque to keep the depth of cut within an acceptable range during the cutting process. To achieve this, machine learning algorithms are applied to build ROP and drilling torque models. Then, a metaheuristic algorithm is used to determine the optimal technical control parameters, the weight on bit (WOB) and revolutions per minute (RPM), that simultaneously enhance the ROP and mitigate excessive vibrations. This paper introduces a new methodology for mitigating drill string vibrations, improving the rate of penetration (ROP), minimizing BHA failures, and reducing drilling costs
  • No Thumbnail Available
    Item
    A machine learning and Particle Swarm Optimization approach for desiccant wheel modeling and performance prediction
    (Elsevier, 2025) Ghersi, Djamal Eddine; Mougari, Nour Elislam; Loubar, Khaled; Amoura, Meriem; Desideri, Umberto
    Accurate modeling of desiccant wheels (DWs) is critical for the design and optimization of energy-efficient dehumidification systems. This study presents a novel approach for predicting DW performance by coupling machine learning (ML) models with Particle Swarm Optimization (PSO) for hyperparameter tuning. To validate the effectiveness of this metaheuristic approach, the performance of the PSO-optimized models was rigorously benchmarked against counterparts tuned using conventional Bayesian Optimization (BO). Four distinct ML models, including Artificial Neural Network (ANN), k-Nearest Neighbors (KNN), Random Forest (RF), and Support Vector Regressor (SVR), were developed to predict the process air outlet temperature (Tp,out) and humidity ratio (ωp,out). The models were trained and validated on a comprehensive dataset, uniquely expanded to include experimental data from low-humidity and low-temperature deep dehumidification conditions. The results demonstrate that the PSO-optimized Artificial Neural Network (PSO-ANN) model provides superior predictive accuracy. For the process outlet temperature, the PSO-ANN model achieved a Coefficient of Determination (R2) of 0.9985 and a Root Mean Square Error (RMSE) of 0.3204 °C. For the outlet humidity ratio, it achieved an R2 of 0.9984 and a RMSE of 0.1497 g/kg. Furthermore, a SHAP (SHapley Additive exPlanations) analysis confirmed that the model’s predictions are physically consistent and interpretable. The developed high-fidelity model serves as a robust and reliable tool for the advanced analysis and design of desiccant air conditioning systems across a wide range of operational scenarios
  • No Thumbnail Available
    Item
    Prediction of Flash Points of Petroleum Middle Distillates Using an Artificial Neural Network Model
    (Pleiades Publishing, 2024) Bedda, Kahina
    An artificial neural network (ANN) model of a multilayer perceptron-type was developed to predict flash points of petroleum middle distillates. The ANN model was designed using 252 experimental data points taken from the literature. The properties of the distillates, namely, specific gravity and distillation temperatures, were the input parameters of the model. The training of the network was carried out using the Levenberg– Marquardt backpropagation algorithm and the early stopping technique. A comparison of the statistical parameters of different networks made it possible to determine the optimal number of neurons in the hidden layer with the best weight and bias values. The network containing nine hidden neurons was selected as the best predictive model. The ANN model as well as the Alqaheem–Riazi’s model was evaluated for the prediction of flash points by a statistical analysis based on the calculation of the mean square error, Pearson correlation coefficient, coefficient of determination, absolute percentage errors, and the mean absolute percentage error. The ANN model provided higher prediction accuracy over a wide distillation range than the Alqaheem–Riazi’s model. The developed ANN model is a reliable and fast tool for the low-cost estimation of flash points of petroleum middle distillates.
  • No Thumbnail Available
    Item
    Prediction of smoke points of kerosene distillates using simple laboratory tests: artificial neural network versus conventional correlations
    (Pleiades Publishing, 2023) Bedda, Kahina
    In the present study, an artificial neural network (ANN) model and three well-known correlations were used to predict the smoke points of 430 kerosene distillates from their specific gravities and distillation temperatures. The ANN model was developed in MATLAB software, it is a feedforward multilayer perceptron with a single hidden layer. The optimal number of neurons in the hidden layer as well as the best training algorithm and the best values of connection weights and biases were determined by trial and error using the nftool command. The early stopping technique by cross-validation was employed to avoid overfitting of the model. The developed model composed of 17 sigmoid hidden neurons and one linear output neuron was trained with the Levenberg-Marquardt backpropagation algorithm. This model allowed the prediction of smoke points with a coefficient of determination of 0.852, an average absolute deviation of 1.4 mm and an average absolute relative deviation of 6%. Statistical analysis of the results indicated that the prediction accuracy of the ANN model is higher than that of the conventional correlations. Indeed, in addition to its effectiveness, the proposed ANN method for the estimation of smoke points has the advantages of low-cost and easy implementation, as it relies on simple laboratory tests. Thus, the developed ANN model is a reliable tool that can be used in petroleum refineries for fast quality control of kerosene distillates.
  • No Thumbnail Available
    Item
    Sensitivity analysis of the gtn damage parameters at different temperature for dynamic fracture propagation in x70 pipeline steel using neural network
    (Gruppo Italiano Frattura, 2021) Abdelmoumin Ouladbrahim, Abdelmoumin; Belaidi, Idir; Khatir, Samir; Magagnini, Erica; Capozucca, Roberto; Wahab, Magd Abdel
    In this paper, the initial and maximum load was studied using the Finite Element Modeling (FEM) analysis during impact testing (CVN) of pipeline X70 steel. The Gurson-Tvergaard-Needleman (GTN) constitutive model has been used to simulate the growth of voids during deformation of pipeline steel at different temperatures. FEM simulations results used to study the sensitivity of the initial and maximum load with GTN parameters values proposed and the variation of temperatures. Finally, the applied artificial neural network (ANN) is used to predict the initial and maximum load for a given set of damage parameters X70 steel at different temperatures, based on the results obtained, the neural network is able to provide a satisfactory approximation of the load initiation and load maximum in impact testing of X70 Steel
  • No Thumbnail Available
    Item
    Artificial Neuron Network Based Faults Detection and Localization in the High Voltage Transmission Lines with Mho Distance Relay
    (IETA, 2020) Boumedine, Mohamed Said; Khodja, Djalal Eddine; Chakroune, Salim
    This study offers the opportunity to extend the functioning of the most advanced protection systems. The faults which can arise on the power transmission lines are numerous and varied: Short-circuit; Overvoltage; Overloads, etc. In the context of short circuits, the conventional sensor as the Mho distance relay also known as the admittance relay is generally used. This relay will be discussed later in this study. By taking into account the preventive risks of the Mho relay and discover the new techniques of artificial intelligence, namely the neural network which can contribute to the precise and rapid detection of all types of short-circuit faults. The results of the simulation tests demonstrate the effectiveness of the methods proposed for the automatic diagnosis of faults.
  • No Thumbnail Available
    Item
    Keratoconus prognosis study for patients with corneal external mechanical stress mode
    (Springer Nature, 2020) Bettahar, Toufik; Rahmoune, Chemseddine; Benazzouz, Djamel
    Purpose To demonstrate the correlation between excessive eye rubbing and corneal degeneration for Keratoconus patients. Materials and methods Keratoconus (KC) patients who regularly rub their eyes had shown a rapid degeneration rate of their affected corneas. This observation is experimentally and numerical discussed and developed based on clinical data of 8 of KC Patients with a mean age of 26.5 ± 9.4 years old, and four healthy individuals with a mean age of 24.33 ± 5 years old at the baseline. Corneal topography was used to measure both central corneal thickness (CCT) and its total refractive power. The registered data had been exploited to assess the progression of the disease, and the final results were embedded in a finite element model of human corneas to simulate their response to eye rubbing at different stages of the pathology. Corneal lifetime prognosis using multi-layer perceptron was then established to estimate the number of eye rubbing cycles for each stage of KC. Results The survey of KC patients who declared stopping eye rubbing had shown a decrease in CCT loss rate, followed by a durable stability. Mechanical stresses numerical simulations had shown different corneal behaviours in term of shape deformity, apical raise and corneal applanation between healthy and KC stages models. Apical rise ranged from 0.122 to 0.389 mm for an applied intraocular pressure that equals to 15 mmHg. A normal stress of 5 kPa provoked a corneal applanation that ranged from 0.27 mm in healthy cases to 1.173 mm in severe stages of the disease. The application of 2.5 kPa biaxial stress had resulted normal and tangential applanations that successively ranged from 0.152 and 0.173 mm in healthy corneas to 0.446 mm and 0.458 mm in severe KC stages. An adopted prognosis algorithm was able to predict the current stage of the disease and to estimate the remaining number of eye rubbing cycles before failure. Conclusion Eye rubbing was proven to be a considerable contributing factor in KC patient’s corneal degeneration. The progression of this pathology could be decreased or halted by stopping eye rubbing at early stages.
  • No Thumbnail Available
    Item
    Application of artificial neural network and kinetic modeling for the prediction of biogas and methane production in anaerobic digestion of several organic wastes
    (Taylor & Francis, 2021) Mougari, Nour El Islam; Largeau, J. F.; Himrane, N.; Hachemi, M.; Tazerout, M.
    In the current study, artificial neural network (ANN) and modified Gompertz equation (MG) were applied to develop integrated based models for the prediction of cumulative biogas and methane yield (CBY and CMY) from anaerobic digestion (AD) of several organic wastes. Volatile solid to total solid ratio (VS/TS), carbon content (C), carbon-to-nitrogen ratio (C/N) and digestion time (DT) were selected as input features for the implementation of ANN approach. Genetic algorithm (GA) was employed in order to optimize the ANN architecture as well as the kinetic parameters of the MG to provide reliable and fast learning for better prediction performance. To evaluate model performances, determination coefficient (R2) and root mean square error (RMSE) were used. Both the approaches performed well in predicting CBY and CMY and showed a good agreement with the experimental data. However, GA-ANN models exhibit smaller deviation and higher predictive accuracy with satisfactory RMSE and R2 of about 0.0045 and 0.9996 for CBY, and 0.0046 and 0.9998 for CMY, compared with GA-MG models. This evinces the effectiveness of the developed approach to forecast CBY and CMY and can be an effective tool for the scale up of anaerobic digestion units and technico-economic studies
  • No Thumbnail Available
    Item
    Predictionofnaturalgashydratesformationusingacombinationofthermodynamicandneuralnetworkmodeling
    (Elsevier, 2019) Rebai, Noura; Hadjadj, Ahmed; Benmounah, Abdelbaki; Abdallah, S.Berrouk; M.Boualleg, Salim
    During the treatment or transport of natural gas, the presence of water, even in very small quantities, can trigger hydrates formation that causes plugging of gas lines and cryogenic exchangers and even irreversible damages to expansion valves, turbo expanders and other key equipment. Hence, the need for a timely control and monitoring of gas hydrate formation conditions is crucial. This work presents a two-legged approach that combines thermodynamics and artificial neural network modeling to enhance the accuracy with which hydrates formation conditions are predicted particularly for gas mixture systems. For the latter, Van der Waals-Platteeuw thermodynamic model proves very inaccurate. To improve the accuracy of its predictions, an additional corrective term has been approximated using a trained network of artificial neurons. The validation of this approach using a database of 4660 data points shows a significant decrease in the overall relative error on the pressure from around 23.75%–3.15%. The approach can be extended for more complicated systems and for the prediction of other thermodynamics properties related to the formation of hydrates
  • No Thumbnail Available
    Item
    Artificial neuron network based faults detection and localization in the high voltage transmission lines with MHO distance relay
    (International Information and Engineering Technology Association, 2020) B.M, Said; K.D., Eddine; C., Salim
    This study offers the opportunity to extend the functioning of the most advanced protection systems. The faults which can arise on the power transmission lines are numerous and varied: Short-circuit; Overvoltage; Overloads, etc. In the context of short circuits, the conventional sensor as the Mho distance relay also known as the admittance relay is generally used. This relay will be discussed later in this study. By taking into account the preventive risks of the Mho relay and discover the new techniques of artificial intelligence, namely the neural network which can contribute to the precise and rapid detection of all types of short-circuit faults. The results of the simulation tests demonstrate the effectiveness of the methods proposed for the automatic diagnosis of faults