ABSTRACT. —In this paper, we present a study on the application of Iterative Learning Control (ILC) to knee exoskeletons in rehabilitation for patients with walking difficulties. Our study explores the effectiveness of ILC in improving trajectory tracking in repetitive tasks such as assisted walking and precise knee positioning. We demonstrate that ILC relies on learning rates to provide stable, precise, and robust control in the face of uncertainties associated with its use. Our simulation results highlight the effectiveness, performance, and robustness of ILC against external disturbances, emphasizing its potential for clinical applications in medical rehabilitation

ABSTRACT. This article focuses on the application of a nonlinear decoupling control PD type on a two-degree-offreedom manipulator robot, which exhibits complex nonlinearities in its modeling. The parameters of the PD controller are determined by considering that the modeling of the Pelican robot is a second-order linear system, with a step response without oscillations. The simulation results on the robot clearly allow us to evaluate the performance of the nonlinear decoupling control in the field of robotics and confirm its effectiveness.

ABSTRACT. Detecting sarcasm in text is a challenging task that has garnered significant interest in the field of natural language processing (NLP). This study presents a comparative analysis of three transformer-based models—BERT, DistilBERT, and RoBERTa—for the task of sarcasm detection. We utilize two datasets: the Sarcasm Headlines Dataset and the ArSarcasm Dataset. Each model is evaluated using standard metrics including accuracy, precision, recall, and F1-score. Our results demonstrate that these models achieve remarkable performance, with accuracy, precision, recall, and F1-scores reaching up to 0.99 on the Sarcasm Headlines Dataset, and scores close to 0.83 on the ArSarcasm Dataset. This comparative analysis highlights the effectiveness of transformer-based models in understanding and identifying sarcastic content in text.

ABSTRACT. In an environment where the spread of information is accelerating exponentially via increasingly accessible social networks, there is a clear and considerable rise in the dissemination of false news, also known as fake news. These have fundamentally altered the landscape of information sharing, with potentially harmful effects on individuals and societies, posing a real threat to the reliability of information. The problem is intensified by the lack of adequate procedures to combat fake news. Consequently, the detection of unfounded fake news has become an urgent issue in media contxt. Although various studies have developed machine learning models to detect fake news in several languages, significant gaps remain in addressing this issue in the Arabic language. This article fills these gaps by analyzing the detection of fake news in Arabic through textual content and their sources, also by using a transfer learning method based on BERT. The study used a large database of over 606,912 news articles covering multiple domains collected from 134 Arabic news websites. Our approach includes innovative feature extraction techniques such as n-grams and document engineering to improve model performance. The results show a significant improvement in accuracy, demonstrating the effectiveness of our approach in identifying fake news in Arabic. This research ultimately contributes to advancing fake news detection in underrepresented languages.

ABSTRACT. This paper is designed to provide a comprehensive overview of the latest developments in fault tolerance methods for cloud computing. Maintaining high availability and reliability of cloud environments requires fault tolerance. This paper explores fault tolerance in the context of cloud computing and discusses recent challenges and innovations in the field. Moreover, it examines the ongoing research efforts to improve fault-tolerance architectures. At the end of the paper, the paper presents system-level metrics that are relevant to fault tolerance

ABSTRACT. Water management is a critical issue affecting agriculture, necessitating efficient water management practices to ensure sustainable crop production. Traditional irrigation methods often lead to overuse or underuse of water, resulting in suboptimal crop growth and wastage of valuable resources. This article proposes an advanced architecture for intelligent irrigation systems that integrates Fuzzy Logic (FL) and Deep Learning (DL). DL algorithms are employed for plant type identification and prediction, and FL is used to estimate the optimal watering duration. The integration of these AI algorithms with sensor networks and real-time data analytics aims to optimize water usage and improve plant yield. The proposed system’s design, implementation, and potential benefits are discussed, demonstrating significant improvements over traditional.

ABSTRACT. In solar energy systems, a solar air heater is a crucial component that is designed to capture solar radiation, convert it to thermal energy at the absorber surface, and then transfer that heat to the air as it passes through the collector. This study evaluates a flat-plate air solar collector's performance, which is intended to improve a solar desalination system that uses humidification and dehumidification procedures. Results that highlight how different operational and design elements affect the collector's efficiency are presented in the report. Significantly, the study shows that thermal efficiency rises with increasing wind speed, peaking at 80% efficiency between 3 and 7 m/s.

ABSTRACT. Abstract - The aim of this paper is to scrutinize the free convection of Al2O3-Ag-CuO/water ternary hybrid nanofluid, with a nanoparticles (NPs) volumetric composition of (30:40:30) respectively, in a differentially vertical heated square enclosure. The governing energy and Navier-Stokes equations are solved using finite volume approach, under several pertinent parameters namely, Rayleigh number (10^3≤Ra≤10^6) and NPs overall volumetric concentration (0%≤φ≤10%). Examination of the influence of flow and thermal fields through isotherms, streamlines, velocity, temperature profiles and Nusselt number is carried out lucidly. Results show that when Ra increases, the convection is intensified and a clockwise vortex develops inside the cavity as a consequences of high-energy transport through the flow associated with the irregular motion of nano-particles. The ternary hybrid nanofluid appears to be the most efficient nanofluid in comparison with the other hybrid nanofluids, particularly at higher total volume fraction, thus it can be used for heat transfer intensification in closed chambers, e.g., for cooling of heat-generating elements or in solar panels applications.

ABSTRACT. Digital transformation is paving a rapid change era. The Industry 4.0 and technological advances such as Artificial Intelligence (AI), manufacturers have powerful tools at their disposal for monitoring and anticipating machine failures. This paper presents an intuitively designed monitoring dashboard aimed at optimizing predictive maintenance. The proposed approach allows operators to observe and respond to real-time predictions, through an intuitive interface that displays the learning model results in an easily accessible way. Bridging the gap between cutting-edge AI research and real-world applications, providing accessible and affordable predictive maintenance for wider operations.

ABSTRACT. This study addresses the modeling and simulation of GaAsPN/GaAs quantum dot solar cell QDSC. This quaternary alloy is one of the III-V semiconductors which gained importance in the recent years for optoelectronic applications due to its perfect lattice matching to Si substrates and to its bandgap that can be decreases drastically with the incorporation of Nitrogen and Phosphorus into GaAs, improving therefore the absorption the wavelengths near the red part. The heterostructures consist of GaAs1-x-yPxNy quantum dots separated by GaAs barrier layers. The width and thickness of the quantum dot are 10 nm and 5 nm, respectively. Our results have been shown that with x=1% and y=2% we obtained optimal efficiency of about 13.96 % using 10 QD layers. However, 30 GaAs0.97P0.01N0.02/GaAs QD layers produce a short circuit current and efficiency of about 19.85 mA/cm2 and 14.3 %, respectively. In addition the same number of QD layers extends the absorption edge of GaAs solar cell from 900 nm to 950 nm with a gain of external quantum efficiency EQE of about 5.487 % at this range.

ABSTRACT. Our work consists in presenting the results of an invention for a membrane distillation system coupled to an efficient and robust water solar collector which produces potable water with high quality and a small percentage of brackish discharge independent of salinity of the water source. Then to optimize and to characterize experimental of the air gap membrane distillation (AGMD) on a pilot plant. During the tests, brackish water was used, ranging from 4.2 to 12.5 g / l of salt. The results show that the permeate flux increases as the temperature and feed rate increase and the thickness of the air gap decreases (from 5.12 to 1.5 mm). Our AGMD system was modeled using Matlab's programming on heat and mass transfer aspects. The 1D model is based on the transfer equations and correlations of the literature present in the membrane distillation pilot. The maximum permeate flux obtained was 7.4 kg / m² h with a hot fluid temperature of 80 ° C., an air gap of 1.5 mm and hot and cold flow rates of 4.8 l / min. For all measurements, the maximum relative difference between the experimental results and the simulated results is observed at 10% errors. The results of low temperature hot fluids can be interesting in the solar energy coupling project.

ABSTRACT. In this work, a nonlinear neural network model based on deep learning was used for the prediction of the geometric accuracy of the Single Point Incremental Forming (SPIF) process. A stacked autoencoders network was chosen, coupled by a backpropagation algorithm to predict the geometric precision of the parts. Geometric accuracy was presented by two responses factors: the springback phenomenon and the thickness variation. However, six parameters were selected as input factors: initial sheet thickness, vertical increment, speed rate, feed rate, tool trajectory, and forming angle. A novel deep learning-based approach was proposed that includes both feature extraction phase and regression phase in a single model to predict geometric accuracy. The proposed approach combines Stacked Auto-Encoders (SAE) with back-propagation algorithm. Finally, the experimental results show that the proposed approach has higher prediction accuracy presented by the correlation coefficient values R2= 0.963 and R2= 0.996 to the springback and thickness variation respectively..

ABSTRACT. The purpose of this work is the development of an adaptive observer in order to enhance the stability of the direct torque controlled permanent magnet synchronous motors. This observer makes it possible to reconstruct the mechanical speed and state variables using stator voltages and currents. Lyapunov theory has been used for the synthesis of this adaptive observer in order to guarantee the closed-loop stability. It is well that the main drawback of the direct torque control is its sensitivity to parametric variation, especially to stator resistance at low speed operation. For this reason, the sliding mode control has been used for the speed correction in order to guarantee control robustness against parametric variation. The undesirable chattering phenomena can be remedied using a smooth continuous function. The contribution of this work is that a new and simple algorithm has been obtained. It has the advantage to be easily implemented in calculators. Simulation tests are provided to evaluate the consistency of this observer at low and high speeds under disturbances. Simulation tests show that the control scheme introduces high performances of robustness, stability and precision under disturbances caused by parametric variation.

ABSTRACT. Object-relational conceptual modelling tools do not support all new concepts available in SQL4. This study presents an enhanced empirical research framework for object-relational database design based on conceptual modelling. This paper describes a proposal for heuristic rules for designing an object-relational database and translating data and relationship semantics from the conceptual schema generated by a UML class diagram to its logical schema counterpart by SQL:2008. A set of rules has been proposed that help visualize and implement object-relational databases, allowing the user to create a script model of a database using SQL. The study explains the importance of these guidelines and gives a detailed explanation of the proposed rules with an illustrative example. The proposed method was evaluated by applying it to a case study and comparing the results with some other methods and proposals. This study confirms that modern technologies and new databases play an important role in the success of software design, so it attempts to facilitate the design process and help software developers understand this type of database and save them time and effort.

ABSTRACT. Software has become very important in everyday life. It is used in almost every field. Even in local society, the dependency level on software has also increased. Whenever there is a software failure, it might have severe consequences on society and individuals alike. Failure does not only affect the client but also the company that is responsible for the development. This study aims to explore the issues that face users and investigate the factors that cause software to fail. A questionnaire was built to derive the issues that users face and distributed using an online Google form across twelve organizations in Benghazi, Libya. In addition, ten interviews were conducted with twenty developers to identify the factors that cause failure. The most common factors among interviewees were identified, including unrealistic schedules and budgets, a lack of user involvement, and incomplete requirements.

ABSTRACT. Radio propagation prediction refers to predicting the behavior of radio signals that travel from one point to another in a vacuum or atmosphere. Since radio signals are a form of electromagnetic waves, they are subjected to various phenomena such as reflection, diffraction, refraction, absorption, scattering, and polarization. It’s important to predict path loss using radio propagation prediction models to calculate the signal strength at the receiver. This study proposes a mathematical propagation model that predicts the path loss at the receiver. The study covered four different location zones within Amara city/ Maysan governorate / Iraq. The data for this study was collected using the Drive Test feature of the TEMS phone, which was connected to the TEMS investigation tool. Based on the results obtained from several empirical propagation models, it was found that the proposed model provided a better match for the data on the measured signal strength. This model showed root mean square error RMSE less than other models, at mobile antenna heights of 1.0m and distances 40, 200, 400, and 600 meters between transmitter and receiver. Other propagation models such as Hata and COST231, which had over-predicted the radio channel, may need further investigation in the future.

ABSTRACT. This paper explores the intersection of mental health and technology, focusing on the detection of personality disorders using social media data and deep learning techniques. Drawing insights from seminal studies, we employ a diverse dataset of 27977 persons to leverage Our findings highlight the potential of linguistic patterns and deep learning in discerning indicators of personality disorders. The study contributes to the ongoing discourse on ethically utilizing technology for mental health, emphasizing the need for continued exploration in diagnostic tools and interventions. Our results are comparative to the best state of the arts methods with an accuracy of 0.89 and recall of 0.95.

ABSTRACT. Detecting bias in news stories is essential to maintaining objectivity and advancing media openness. Studying languages like Amharic, which have little resources and research in natural language processing, makes the task more difficult. In this study, we proposed a novel method that uses deep learning techniques to identify bias in Amharic news articles. To tackle this task, we gathered a broad dataset of Amharic news stories from online social media, covering a range of subjects and viewpoints. Experts in the field accurately annotated the dataset, classifying each piece as neutral or biased according to the underlying story. Our deep learning model is trained and evaluated using this annotated dataset. Convolutional neural network (CNN) architecture was utilized to extract the linguistic patterns and contextual information from the Amharic news text. Using the labeled dataset as a training set, the model was optimized to detect bias with high accuracy and precision. We optimized the model's performance by fine-tuning its hyperparameters through considerable experimentation. Our model's efficacy in identifying bias in Amharic news items was evaluated using a different test set, and the findings proved. With a high degree of confidence, our model was able to identify biased content with an accuracy of 89.5% and a f1-score of 80.28%. This study revealed that it improves media literacy, encourages objective reporting, and makes it easier for the Amharic-speaking community to verify information.

ABSTRACT. This study aimed to develop and evaluate an Android-based mobile application relative to mental health services in Butuan City to address challenges faced by both clients seeking mental health services, and mental health service providers who are also trying to reach clients and patients in need of their services. This mobile application functions as an intermediary platform, allowing clients to conveniently search for mental health service providers, schedule appointments, and locate their physical office addresses. For mental health service providers, the application offers appointment management modules and service management, streamlining their workflow and potentially reducing administrative burdens such as scheduling and other workloads. To evaluate the effectiveness of the mobile application in addressing the identified challenges, a survey was conducted among 100 respondents within Butuan City, including potential clients and mental health service providers. The survey results indicate that the application meets its intended goals of improving access to mental health services and streamlining appointment management for service providers. An evaluation of the application’s usability was also conducted using the System Usability Scale (SUS). The survey results yielded an overall weighted mean of 84.625, translating to an adjective rating of "Excellent." This high score indicates a strong level of agreement among the respondents regarding the application's usability. These findings suggest that users perceive that the mobile application is a user-friendly and intuitive platform. Qualitative feedback obtained during the usability study highlighted the application's value in addressing current challenges faced by both clients and mental health service providers.

ABSTRACT. In this paper an examination has been performed for the steady-state voltage stability using different FACTS devices. The simulation of IEEE 14 Bus power system including FACTS devices has been carried out in the PSAT/MATLAB domain. The performance of unified power flow controller (UPFC) for the upgradation of transient stability limit has been investigated along with other FACTS devices such as Thyristor Controlled Series Compensator (TCSC), static synchronous series compensator (SSSC), STATIC VAR compensator (SVC) and static synchronous compensator (STATCOM). The document establishes the profitability and performance of UPFC on other FACTS.