Abstract The rapid growth of electric vehicle (EV) adoption demands intelligent, secure, and resilient charging infrastructures capable of operating reliably under cyber-physical uncertainties. This paper designs a Composite Vulnerability Index (CVI)-based mitigation-aware control law for real-time, resilient management of multiple EV charging systems. The CVI quantifies the aggregated risk level of each EV’s charging context by fusing multiple factors such as cyber threats, communication latency, thermal stress, and power quality disturbances. The paper proposes a Digital-Twin enhanced Reinforcement Learning (RL) framework that integrates a Composite Vulnerability Index (CVI) with a mitigation-aware control law for large-scale multi-EV charging networks. A physics-informed EV Charging Digital Twin (DT) continuously predicts system evolution and identifies discrepancies using a learned residual model, enabling real-time characterization of operational uncertainty and cyber-attack impacts. The DT outputs both corrected state estimates and a risk cost that penalizes model uncertainty and elevated vulnerability levels. A Dreamer-V3 world-model-based RL agent uses these DT signals to learn an optimal control policy that adaptively allocates charging power, mitigates risks, and preserves grid safety under DDoS, MITM, spoofing, voltage instability, thermal overload, and communication delays. The proposed CVI integrates cyber, physical, and communication vulnerabilities into a unified state-aware metric, enabling coordinated power redistribution among multiple EV stations through a distributed primal-dual scheme. Simulation results with distributed EV charging stations demonstrate the system’s ability to ensure adaptive charging, equitable energy distribution, and successful convergence to target state-of-charge (SOC) levels while mitigating potential threats. The findings validate the CVI-based strategy as an effective framework for secure, scalable, and risk-aware EV energy management in next-generation smart grids.

Abstract Sequential Orthogonal Frequency Division Multiplexing (SeqOFDM) is a modulation scheme riding on OFDM and shown to provide a better spectral efficiency without introducing undesirable degradation in the Bit Error Rate (BER). In the scheme, a mapper is added at the input of an OFDM transmitter. This mapper is a lookup table that helps select three carriers for every four bits (or symbols) presented. These carriers form the OFDM subcarriers for signal transmission. At the receiver output is a demapper that maps the received carriers to the table of the equivalent bits effectively doing the opposite of the mapper. This paper presents a theoretical analysis of the scheme, deriving closed-form expressions for BER over Additive White Gaussian Noise (AWGN) channel. Theoretical results are benchmarked against Monte Carlo simulations under identical configurations, with findings demonstrating a strong agreement between theoretical predictions and simulated outcomes, particularly in the mid-to-high Signal to Noise Ratio (SNR) regime, confirming the accuracy of the derived model. An extension of the derivation is made to Rician and Rayleigh channels, The results offer a reliable foundation for the design and optimization of SeqOFDM communication systems which at the same time circumvents the computationally intensive and time-consuming process of BER estimation through Monte Carlo methods. Thus, the objective of providing a fast and scalable alternative for performance evaluation is met.

Abstract This paper introduces a novel methodology for determining the photovoltaic (PV) hosting capacity of distribution networks by integrating short-circuit impedance analysis with maximum PV penetration assessment under dynamic stability, inverter control, and grid code considerations. This methodology utilizes Power System Simulation for Engineering (PSS/E) software by applying multiple equivalent Thevenin impedance values to a typical distribution network. The findings of this methodology are then applied to a part of Egypt’s distribution network to validate the obtained results and find the maximum hosting capacity. The second objective is to increase the PV hosting capacity by applying different control techniques to PV inverters. Results show that the PV inverter can provide voltage support and act as an active source in the power system to achieve the grid code requirements. It is also concluded that we can predict the hosting capacity of any bus in a distribution network. The results obtained from the typical system are in good agreement with those obtained from the conventional analytical method applied to the fundamental part of the Egyptian distribution network. Based on static and dynamic assessments, this research helps utility operators make decisions about large-scale PV integration into the distribution network.

Abstract Malnutrition and food insecurity have remained critical issues faced in Africa. According to the 2023 statistical data in Nigeria, for instance, approximately 87 million out of the country’s 220 million people (39.5 per cent) still live below the poverty line. From this trajectory, it is safe to say that the continent of Africa is not yet on the path to Zero Hunger (SDG 2) by 2030. While several successive government administrations have proposed various intervention programs such as Operation Feed the Nation, the Green Revolution, and Lower Niger River Basin Development Authority, fertiliser support, among others, in Nigeria, little attention has been given to plant diseases, one of the root causes of low productivity among smallholder farmers. Therefore, this research leveraged the inherent characteristics of deep learning models and developed an improved deep learning Plant Doctor based on MobileNetV3-Small architecture with a user-friendly interface that enables the drag and drop of plant images or direct upload. The developed system was tailored towards the computational demands of smallholder farmers’ low computing devices. The developed MobileNetV3-Small architecture uses a smart patch-based scanning process that focuses on the leaf regions, resizes the image to 224 × 224 as input size, and unfreezes 20 layers for feature learning from the patches. The patches are augmented using brightness, contrast, and slight rotation to ease the detection of tiny symptoms. This allows for detailed symptom analysis without overwhelming memory or processing power. The developed, improved MobileNetV3-based plant doctor easily detects plant diseases through their visual symptoms on their leaves, prescribes treatments, and broadcasts detected diseases to farmers within the same region for preventive control measures. The evaluation of the developed MobileNetV3-small showed that the system can detect plant disease with an accuracy of 99.85% on the merged PlantVillageDoc dataset, with the added advantage of broadcasting detected plant diseases to other farmers within the same cluster through their registered email. This system offers a paradigm shift in educating smallholder farmers by providing timely disease detection and expert guidance, thereby reducing crop losses, improving yields, and strengthening national food security.

Abstract This research paper focuses on the design, implementation, and evaluation of a digital amplifier system, engineered to incorporate MOSFETs and H-bridge configurations. The primary function of this digital amplifier is to super-impose alternating voltage onto a direct current (DC) power supply. A key aspect of the study highlights the strategic implementation of a Second-Order Generalized Integrator (SOGI) model. This SOGI model is seamlessly integrated within the digital amplifier’s control framework to precisely generate and maintain the desired voltage waveform characteristics. The primary objective of super-imposing this alternating output voltage onto the DC power supply is to facilitate a comprehensive evaluation of a voltage class B system’s immunity to DC-side ripple. By subjecting the system to this super-imposed voltage, the research aims to rigorously assess its ability to maintain reliable performance even when subjected to fluctuations and disturbances on the DC power supply line. This setup enables testing and assessment of the electrical performance, safety, and durability of voltage Class B systems and components. The digital amplifier system was extensively tested and validated under both low voltage (4 V) and high voltage (400 V) conditions. The results demonstrate the system’s capability to generate a sinusoidal output voltage with a magnitude of up to 40 V, operating across a wide frequency range from 10 Hz to 150 kHz. The incorporation of the SOGI model enabled effective processing of single-phase AC signals and permits the real-time adjustment of frequencies. The system successfully super-imposed the generated sinusoidal voltage onto the main DC power supply, delivering the combined signal to the load. The amplifier’s performance was validated under high voltage conditions (up to 400 V DC input), showcasing its stability and reliability across diverse operating parameters. This work lays the foundation for advancing power electronic technologies and enabling their deployment in a wide range of industrial applications.

Abstract Efficient and accurate energy consumption modeling is fundamental for optimizing grid stability, forecasting demand and developing sustainable energy policies. This paper introduces the Shifted-Mode Lomax (SM Lomax) regression model, a novel statistical framework designed to address two key challenges in energy datasets: non-zero peaks reflecting baseline consumption and heavy-tailed distributions that capture extreme usage events. Traditional models, such as the Gamma and standard Lomax distributions, often fail to accommodate these features simultaneously due to the unrealistic zero-mode assumptions. Rooted in weighted distribution theory, the SM Lomax model inherently supports positive modes while providing enhanced control over tail behavior. Its regression framework connects location and shape parameters to covariates, enabling interpretable predictions of peak demand. Simulations confirm the asymptotic unbiasedness and consistency of the MLEs. Empirical validation using 105,140 smart meter readings from the Tanzania Electric Supply Company (TANESCO) demonstrates a 22.8% improvement in the Akaike Information Criterion (AIC) over benchmark models.

Abstract The increasing popularity of multimedia applications, such as video classification, has underscored the need for efficient methods to manage and categorize vast video datasets. Video classification simplifies video categorization, enhancing searchability and retrieval by leveraging distinctive features extracted from textual, audio, and visual components. This paper introduces an automated video recognition system that classifies video content based on motion types (low, medium, and high) derived from visual component characteristics. The proposed system utilizes advanced artificial intelligence techniques with four feature extraction methods; MFCC alone, (2) MFCC after applying DWT, (3) denoised MFCC, and (4) MFCC after applying denoised DWT. And seven classification algorithms to optimize accuracy. A novel aspect of this study is the application of Mel Frequency Cepstral Coefficients (MFCC) to extract features from the video domain rather than their traditional use in audio processing, demonstrating the effectiveness of MFCC for video classification. Seven classification techniques, including K-Nearest Neighbors (KNN), Radial Basis Function Support Vector Machines (SVM-RBF), Parzen Window Method, Neighborhood Components Analysis (NCA), Multinomial Logistic Regression (ML), Linear Support Vector Machines (SVM Linear), and Decision Trees (DT), are evaluated to establish a robust classification framework. Experimental results indicate that this denoising-enhanced system significantly improves classification accuracy, providing a comprehensive framework for future applications in multimedia management and other fields.

Abstract Pneumonia is a major cause of death worldwide, with nearly 2.2 million deaths every year, including over 700,000 children aged five years and below. Chest X-ray (CXR) imaging is the standard method taken up to diagnose pneumonia; however, such image inspection is shown to be difficult even for expert radiologists. The intricacy in the visual patterns generated in X-ray images often results in misdiagnosis, indicating the importance of efficient and accurate automated substitutes. In this paper, we present a new machine learning-based system that incorporates deep learning combined with handcrafted feature extraction techniques for sophisticated pneumonia detection. We use ResNet-50 for deep feature extraction, with the integration of 2D-discrete wavelet transform (DWT) and Gray-Level Co-occurrence Matrix (GLCM) for texture feature extraction, for the intention of gaining helpful spatial and frequency-domain features. The features obtained are inputted into a Support Vector Machine (SVM) classifier, which is optimized for high accuracy and robust prediction. The experimental findings indicate that the proposed model produces a classification accuracy of 97%, accompanied by an F1-score of 0.97, over traditional methods. Through the synergistic integration of handcrafted and deep learning-based feature extraction techniques, our approach presents a trustworthy and efficient solution for automated pneumonia detection. The proposed method has the potential to aid radiologists in providing timely and accurate diagnoses, thus enhancing patient outcomes and curtailing the global burden of pneumonia-related mortality.