Journal of Advanced Research in Applied Artificial Intelligence and Neural Network

AI-Enhanced Fault Mitigation in Solar-Wind Hybrid Systems: A Review

2026-01-28T05:47:06+00:00

Interpretation of the above statement: Currently, with the increasing penetration of renewable energy sources into power grids, fault detection, fault mitigation, and system reliability have become serious issues with solar and wind hybrid systems. Hence, hybrid renewable energy systems (HRESs) may face several faults, such as mismatch, line-to-line, arc, and ground faults in PV systems, besides the infantness and variability of wind systems. Faults in HRES threaten the reliability of the power supply and thus the stability and safety of equipment in the grid. Relay protection and circuit breakers under dynamic grid conditions could be slow, thereby not responding optimally. Recent advances in fault current limiting using superconductors, including hybrid current limiters, have been developed, characterised by fast response and excellent fault current reduction. Also, AI approaches have been employed to optimise relay coordination and improve transient stability, including, but not limited to, PSO, ACO, and DRL. This review acts as a comprehensive state-of-the-art survey on fault detection and removal techniques within the solar-wind hybrid system, aligns their performance, discusses implementation issues, and their possible integration with smart grids. Comparative results indicate that SFCL can reduce the fault current by almost 70%, but AI-coordinated relay protection drastically improves response time. Cost, scalability, and real-time adaptability remain worthy concerns for the mass deployment of these technologies, while the paper finally concludes that ensuring the reliability, efficiency, and resilience of future renewable energy-based grids shall depend on AI-supported fault current limiters.

A Comprehensive Review and Taxonomy on Machine Learning and Deep Learning Approaches for Brain Tumor Classification

2026-05-04T11:25:28+00:00

Cancer happens to be one of the most lethal diseases, resulting in high mortality rates. An unprecedented global surge in cancer cases can be seen, which leads to the necessity of developing newer methods for cancer detection so as to detect and arrest the progression of the disease quickly. Out of all cancer types, brain cancer happens to have a relatively higher mortality rate. Moreover, the brain being an internal organ makes invasive examinations challenging, thereby relying more on image-based prognosis. With an increased number of cases and extensive pressure on the current medical infrastructure, machine learning and deep learning (ML & DL)-based approaches are being explored extensively to automate some of the procedures so as to aid the medical practitioners by providing useful insights from medical data. In context to brain cancers, one of the foremost steps is identifying brain tumours, segmenting them and potentially classifying them as benign or malignant. The divergences in texture and sites of the tumours make this process complex and challenging to yield high accuracy. This paper presents a comprehensive review of the various contemporary image pre-processing, feature extraction and classification techniques employed in current literature to lay a foundation for future research in the domain with the objective of attaining high classification accuracy.

DOI: https://doi.org/10.24321/3117.4787.202601

Deep Learning Techniques for Cybersecurity in Critical Infrastructure

2026-05-04T11:30:26+00:00

The part of AI-powered cybersecurity is an amazing solution for securing complex infrastructure. The cure for augmenting or increasing the resilience of infrastructure such as power grids, water treatment plants, and transportation systems includes automated threat detection and incident response, along with prevention by AI. Moreover, AI-based threat hunting predicts unseen threats to help security teams prevent risk events from developing into major issues. Yet, significant challenges such as data quality, model biases, explanations, and skills shortages must be overcome to unlock the full potential of AI in cybersecurity. Finding the balance between a lively electronic future for critical infrastructure and its protection requires her to spur more research and development in adversarial machine learning.

DOI: https://doi.org/10.24321/3117.4787.202603

Anomaly detection in human behaviour using computer vision: A review

2026-05-04T12:20:37+00:00

To improve the safety of the public, hundreds and hundreds of CCTV cameras are being installed in public spaces such as roads, shopping centres, parks, etc., but security organisations’ capacity for monitoring cannot keep up. Monitoring relies on human judgement, which is affected by factors such as distraction and stress. Moreover, manual monitoring also comes with the added risk of missing an anomalous event or a delay in reporting. All of this contributes to the need for automation in surveillance. A robust machine learning model that uses computer vision techniques, such as image and video data analysis, object detection, and motion tracking, can help mitigate these problems. Starting with introducing supervised learning methods like Convolutional Neural Networks (CNNs) to learn normal and anomalous data patterns and using Long-Short Term Memory (LSTMs) to capture and analyse time-related dependencies. Moving on to autoencoders, it is an unsupervised learning approach that significantly improves anomaly detection by overcoming the need for a detailed labelled dataset. They learn to reconstruct the normal patterns during training, and during inference, any deviations are marked as an anomaly. Such an autonomous system will allow organisations to focus on tasks that require human attention and help minimise the response time of authorities in the event of anomalous events. This paper presents a comprehensive review of the research done on this topic.

DOI: https://doi.org/10.24321/3117.4787.202602