Belgacem Bouallegue

The proliferation of Internet of Things (IoT) platforms in precision agriculture enables unprecedented resource optimization and operational efficiency, yet it exacerbates critical challenges in sustainable energy management and long‐term infrastructure viability. Accurate forecasting of IoT device energy consumption is essential to extend battery life, reduce maintenance costs, and ensure reliable operation in remote agricultural settings. While deep learning models have shown promise in energy prediction, existing approaches are often narrowly technical, offering limited interpretability and actionable guidance for system optimization. This study bridges this gap by introducing a novel hybrid transformer–long short‐term memory (LSTM) model enhanced with explainable artificial intelligence (XAI) to not only improve prediction accuracy but also deliver actionable engineering insights for energy‐efficient IoT deployment. Our proposed hybrid architecture uniquely combines the transformer’s ability to capture long‐range temporal dependencies with the LSTM’s proficiency in modeling local sequential patterns, effectively addressing the multiscale energy dynamics typical of agricultural IoT operations. Evaluated on a comprehensive real‐world dataset from ESP32‐based agricultural monitoring nodes, the model achieves a mean absolute error (MAE) of 0.2666, outperforming standalone LSTM (MAE: 0.2929) and transformer (MAE: 0.2740) benchmarks. More significantly, SHapley Additive exPlanations (SHAP)–based interpretability analysis identifies idle current and transmission parameters as the dominant factors influencing energy consumption, providing clear optimization targets for system designers. A key empirical finding demonstrated through rigorous XAI is that the choice of programming model, including the use of real‐time operating systems (RTOS), has a negligible impact on overall energy use. This insight liberates engineers from a longstanding design trade‐off, permitting the adoption of advanced RTOS for enhanced reliability and real‐time performance without incurring significant energy penalties. This research contributes to the energy systems community by (1) advancing hybrid deep learning architectures for multiscale energy forecasting; (2) providing interpretable, data‐driven insights for energy optimization in IoT‐enabled agriculture; and (3) offering empirical evidence that redefines design constraints in low‐power embedded systems. The proposed framework not only enhances the longevity and reliability of agricultural IoT networks but also informs energy policy and planning for sustainable digital agriculture. By integrating predictive analytics with strategic governance principles, this work provides a decision‐support framework that enables farm managers to make evidence‐based choices regarding technology adoption, maintenance scheduling, and resource allocation, ultimately supporting the transition toward sustainable, data‐driven agricultural practices.

Breast cancer is the primary cause of death for women around the world, necessitating the development of highly accurate, interpreted, and technologically advanced predictive approaches to support early diagnosis and treatment. In this research, we introduce a deep learning (DL) model for predicting breast cancer using both public and private datasets. The model uses the internet of things (IoT) to improve data collection and real‐time monitoring, and it also uses the SMOTE method to resolve issues of class imbalance. The proposed model combines an explainable AI approach with SHAP values to ensure model interpretability. To identify the best DL algorithm for this method, we assess and compare six different DL algorithms: temporal convolutional networks (TCNs), neural factorization machines (NFMs), long short–term memory (LSTM) networks, recurrent neural networks (RNNs), gated recurrent units (GRUs), and deep kernel learning (DKL). IoT devices allow for the continuous acquisition of patient data, which, when integrated with our predictive models, improve the capacity for early detection. Reliable cancer detection relies on our method’s enhanced predictive accuracy and sensitivity. Furthermore, we offer crucial transparency in clinical settings by using SHAP to give detailed explanations of model decisions. By employing thorough statistical analysis and cross‐validation, we guarantee that our model is resilient and can be applied to various patient populations. The results show that our proposed IoT integrated method has the potential to improve prediction performance and boost confidence in AI‐powered medical diagnostics by making them more accessible and easier to use. From a performance perspective, the proposed approach, which uses the TCN algorithm and SMOTE, achieved the best accuracy for BC prediction. With the public dataset, the experimental results were 99.44%, 100.0%, 99.01%, 98.75%, 99.37%, and 99.89% for accuracy, sensitivity, specificity, precision, F1‐score, and AUC, respectively. The experimental results for accuracy, sensitivity, specificity, precision, F1‐score, and AUC using the private dataset were 97.33%, 93.33%, 100%, 100%, 96.55%, and 99.48%, respectively. On the other hand, with the combined datasets, the TCN algorithm achieved 100% for all performance metrics.

Plant diseases imperil global food security, decimating crop yields and endangering farmers’ livelihoods. Rapid, accurate detection remains a challenge, particularly in resource‐constrained environments lacking portable tools. Our contribution, Agri Bot, introduces a pioneering deep convolutional neural network (CNN) model, uniquely optimized for mobile deployment, transforming plant disease diagnosis. This novel model integrates a lightweight architecture with advanced feature extraction, achieving an exceptional 97.30% accuracy and 98.76% area under the curve (AUC). Unlike computationally intensive traditional CNNs, Agri Bot’s innovative design—featuring a hybrid convolutional autoencoder, max pooling, and dropout layers—ensures high‐speed, real‐time performance on mobile devices. Comparative studies reveal Agri Bot’s superiority, surpassing state‐of‐the‐art models like VGG16 (71.48%) and ResNet50 (96.46%), while rivaling InceptionV3 (99.07%) with significantly lower computational demands. By delivering precise, accessible diagnostics to remote regions, Agri Bot revolutionizes agricultural disease management, enhancing crop resilience and global food security.

The improved virtual flux–based direct power control (VF‐DPC) strategy for a grid‐connected power converters is proposed in this paper to eliminate current harmonics and mitigate voltage fluctuations during grid voltage faults without using voltage sensors. The proposed control strategy integrates VF‐DPC for static reactive power compensator (STATCOM) and active power filter (APF) into a single algorithm. The reference electrical quantities for each operating mode of the converter are selected based on the state of the electrical grid. Specifically, if the grid voltage variation remains within the limits defined by international standards, the APF operating mode is prioritized. However, if the voltage variation exceeds these standards, the STATCOM operating mode is given preference. The primary goal of the proposed control strategy is to enable a seamless transition between APF mode and STATCOM mode, while minimizing reactive power exchange with the electrical grid and reduces the number of sensors required. This control strategy gives priority to the reactive power injection during grid voltage faults to support voltage regulation. Under normal grid conditions, it enables the power converter to effectively eliminate current harmonics produced by nonlinear loads. Simulation results demonstrate that the proposed modified VF‐DPC technique has high performance in both of converter operation modes. Practical results plotted using a test bench based on a dSPACE 1104 card justify the validity and efficiency of the developed algorithm.

Many traditional distribution systems (TDSs) flaws and weaknesses are fixed when renewable distributed generations (RDGs) is integrated into them. Some of the forces that have driven work on the integration of renewable sources into modern conventional power systems are the effective strategies for increasing system efficiency and reducing total cost. This paper introduces the particle swarm optimization (PSO) method for solving problems related to optimal power flow (OPF) that involve solar photovoltaics (PVs) and wind turbines (WT). The aim is to improve the algorithm’s ability to conduct comprehensive searches for the best possible solution. Therefore, a modified investigation into the impact of such integration on power losses and cost reduction at a large conventional power grid approach using PSO to choose the hourly best load flow in the 118 bus IEEE system and utilize MATPOWER for power grid simulations for communication network modeling with RDG integration under various operational situations. Simulation results confirmed that the algorithm can be an efficient choice to solve the OPF problem, minimize the number of generators (Gs), power losses, and cost compared to the fuel source. This provides a deep analysis of how to combine the benefits of solar and wind power to increase the sustainability and economics of a power grid, with salient conclusions for the energy industry aiming to improve grid performance.

In the trajectory of carbon dioxide (CO ₂ ) neutrality, wind and solar energies will be the key for the energy transition in the electricity sector; however, a massive integration of solar and wind farms into the electricity grid by 2050 will be carried out. For this end, powers control and powers management of these two renewable energies have taken the attention of several researchers since the last decades. This article presents a reactive energy management strategy for a power grid linked to a wind farm utilizing doubly fed induction generators (DFIG) and enhanced by a static reactive power compensator (STATCOM). This management strategy improves an electrical grid capability in the event of a low‐voltage ride through (LVRT) and aims to optimize the sizing of the STATCOM to be installed alongside the wind farm. The proposed oriented voltage control strategy (VOC) for the grid‐side converter and STATCOM facilitates effective reactive current injection into the grid during symmetrical faults with significant voltage sags. A maximum power point tracking (MPPT) approach combined with stator flux‐oriented control (FOC) applied to the rotor side converter enables effective control of the DFIG during an asymmetrical fault. Breaking down currents and voltages into positive and negative sequences expressed in the synchronous frame enhances DFIG and STATCOM control during grid voltage asymmetrical faults. The control algorithms are validated by simulation results using MATLAB‐SIMULINK.

Deaf and dumb people struggle with communicating on a day-to-day basis. Current advancements in artificial intelligence (AI) have allowed this communication barrier to be removed. A letter recognition system for Arabic sign language (ArSL) has been developed as a result of this effort. The deep convolutional neural network (CNN) structure is used by the ArSL recognition system in order to process depth data and to improve the ability for hearing-impaired to communicate with others. In the proposed model, letters of the hand-sign alphabet and the Arabic alphabet would be recognized and identified automatically based on user input. The proposed model should be able to identify ArSL with a rate of accuracy of 97.1%. In order to test our approach, we carried out a comparative study and discovered that it is able to differentiate between static indications with a higher level of accuracy than prior studies had achieved using the same dataset.

Deaf and dumb Muslims face educational barriers. They can't read, recite, or comprehend the Holy Qur'an, hence they can't practise Islamic ceremonies. This study proposes a CNN-based Qur'anic sign language recognition methodology. First, photos are used to train for dynamic and static gesture recognition. Second, preparing images diversifies datasets. Finally, CNN-based deep learning models extract and classify features. To teach the deaf and dumb Islamic ceremonies, the programme recognises Arabic sign language hand motions referring to dashed Qur'anic letters. Only 24,137 photos of the Holy Qur'an's 14 dashed letters were used in the trials from ArSL2018, a huge Arabic sign language collection. SMOTE raises training and testing accuracy to 98.31% and 97.67%, respectively, whereas the proposed model reaches 98.05% and 97.13%. RMU obtains 98.66% and 97.52% training and testing accuracy, whereas RMO achieves 98.37% and 97.36%.