Velammal College of Engineering and Technology

Stego-malware in compressed audio poses a covert threat by hiding commands in imperceptible regions, bypassing traditional detection methods. Hence, a novel, Harmonic-Aware Bidirectional Attention Network is proposed to overcome these issues and classify normal audio from stego audio. Initially, existing feature extractors suffer from quefrency aliasing and the loss of fine-grained phase and harmonic details, making them inadequate for identifying subtle steganographic cues in polyphonic environments. Thus, Harmonic-Chroma-Q Transform Feature Extractor (HCQT-FE) is introduced, which uses Harmonic-Percussive Source Separation (HPSS) for isolating tonal elements, Mel-Chroma-Spectral Contrast for enriching spectral and harmonic representations, and Constant-Q Transform (CQT) to preserve log-scaled frequency details aligned with audio structure, thereby capturing multi-dimensional audio representations. Moreover. traditional deep learning models struggle with polyphonic audio due to their inability to capture long-range dependencies amid overlapping harmonic structures, asynchronous onsets, and fluctuating tonal centers. Therefore, a Bidirectional Gated Attention Network (Bi-GAttNet) is presented, which uses Bidirectional LSTMs to model temporal context in both directions and a Gated Attention mechanism to focus on salient harmonic regions while suppressing noise dynamically enhances the ability to detect and classify covert steganographic patterns embedded within audio, achieving improved accuracy, recall, and robustness in stego-malware detection tasks.

Background: Catechol (CC) is an important phenolic molecule that plays an important role in various industrial applications. At the same time, it is a significant environmental pollutant. The rapid, sensitive, and selective detection of CC remains challenging because of the coexistence of structurally similar phenolic impurities. This study proposes an electrochemical method for CC detection using LaBi2O4 and LaBi2O4/reduced graphene oxide (LaBi2O4@rGO) nanocomposites. Methods: The nanocomposites were synthesized using a simple hydrothermal technique, which led to improved electrochemical properties owing to the synergistic effect between LaBi2O4 and rGO. The structural and morphological properties of the LaBi2O4@rGO nanocomposite were investigated using various techniques. Electrochemical analyses were performed using cyclic voltammetry and differential pulse voltammetry, which revealed considerable sensitivity and selectivity for CC detection. Significant findings: The LaBi2O4@rGO incorporated GC electrode possessed a wide linear detection range (1-700 & micro;M), with a low detection limit (0.15 & micro;M). Moreover, the newly fabricated electrode exhibited high stability, repeatability, and reproducibility, making it suitable for environmental applications. Its effectiveness was further validated by detecting CC in real water samples, such as tap and river water, with a 99% recovery rate, thereby demonstrating its immense potential for practical environmental monitoring.

The rapid digitalization of energy systems through smart meters provides vast data streams that, when contextualized with socioeconomic information, can reveal the underlying behavioral mechanisms of electricity use. However, standard machine learning pipelines often overlook the non-Euclidean relationships embedded within social structures and spatial proximity, leading to limited predictive reliability. This paper presents a Hypergraph-Driven Hybrid Intelligence Framework (HHIF) that synthesizes hypergraph attention, graph convolution, and temporal learning to create an adaptive model for residential load forecasting. The proposed system encodes complex socio-energy interactions using Hypergraph Attention Networks (HANs), integrates K-means clustering for structure-aware feature extraction, and fuses GCN and LSTM modules for joint spatial–temporal optimization. Experiments on multi-city smart meter datasets highlight substantial gains in accuracy, stability under noise, and interpretability, underscoring the framework’s applicability for grid management, peak load mitigation, and energy equity research.

Microgrids based on renewable energy have gained prominence as environmental sustainability has received more attention globally. However, the fluctuating nature of renewable energy sources (RES), erratic energy consumption, and the dynamic influence of EVs on microgrid security pose a threat to their stability. This manuscript proposes a hybrid energy management strategy (EMS) for a multi-source microgrid that incorporates electric vehicles (EVs), RES, and an adaptive fractional order proportional-integral-derivative (AFOPID) controller to address these issues. The proposed approach combines the Portia Spider Algorithm (PSA) and Temporal Dynamic Graph Neural Network (TDGNN), which is termed the PSA-TDGNN method. The primary goal of the proposed method is to develop system efficiency, minimize grid dependency, minimize estimation error, and maintain a stable and reliable power supply for connected loads. The TDGNN is used to predict performance under high EV penetration levels, while the PSA is employed to optimize the AFOPID controller's gain factors. The proposed technique is assessed and compared utilizing the MATLAB platform to other existing approaches. The proposed approach yields better outcomes compared to existing methods like the Advanced-Sine Cosine Algorithm (a-SCA), Adaptive Salp Swarm Algorithm (ASSA), and Bald Eagle Search (BES). The proposed approach achieves an error of 0.59%, efficiency of 97%, and computational complexity of 18.23 s. These results demonstrate that the proposed technique achieves better performance and enhances energy management in multi-source microgrids based on EVs compared to existing methods.

Developing efficient and cost-effective catalysts for industrial-scale hydrogen (H2) and oxygen (O2) production is crucial for the clean energy transition. Atomically dispersed single-atom catalysts (AD SACs) (e.g., iron (Fe), cobalt (Co), and nickel (Ni)) without precious metals are particularly promising, offering 100% atomic utilization and optimum catalytic efficiency. Nonetheless, key challenges remain, including the suppression of atom aggregation, the optimization of support structures, and the maintenance of stability. Moreover, industrial catalysis demonstrates that active components with suitable supports are critical, and catalytic efficacy often increases as the size of the active components decreases, opening new frontiers in catalyst design. Thus, metal-organic frameworks (MOFs) and MXenes have emerged as effective anchoring matrices, enabling modulation of electronic structures, rapid charge transfer, and enhanced stabilization of single metal atoms. This review highlights recent progress in the development of AD SACs anchored on MOFs and MXenes, emphasizing the fundamental parameters governing hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) performance. Furthermore, we integrate theoretical insights from Density Functional Theory (DFT) with advanced operando and in-situ characterization techniques to provide a comprehensive understanding of the mechanistic origins governing HER/OER activity. We anticipate that these insights will contribute to improving water-splitting efficiency and sustainability, which are vital for advancing the clean energy revolution. Finally, this review outlines future directions for the development of high-efficiency, cost-effective MXene- and MOF-based SAC catalysts, underscoring their potential in clean energy production.