Southern Journal of Research

Development of Generalized Refinement Strategies in Composite Stationary Iterative Solvers for Linear Systems

Maheen Kanwal — 2025-07-24

Linear equations have many applications in natural sciences, engineering, business, social sciences, and medicine. However, solving these types of systems is an important challenge in Numerical Linear Algebra (NLA). There are two types of Methods to solve the systems: direct approaches and indirect approaches. Iterative methods are very successful in solving large, sparse linear problems. Iterative approaches often outperform direct methods, particularly when working with sparse coefficient matrices.

This research introduces the composite stationary iterative approach for solving linear systems of equations (GCST). And validate by comparing the proposed method with existing methods, in terms of spectral radius, no. of iterations, and convergence rate. Under specific conditions, proposed methods efficiently solve linear systems with coefficient matrices that are irreducibly diagonally dominant (IDD), strictly diagonally dominant (SDD), M-matrices, or symmetric positive definite. MATLAB (R2014b) was used to compute numerical tests.

Numerical Study of Suction Injection Driven Nanofluid Through Parallel Plates Channel

Hina Bashir — 2025-07-24

The main purpose of this research is to analyze the effects of nanofluids that are based on the injection and suction parameters, where the channel is structured using parallel plates. For making governing PDEs dimensionless, we use similarity transformations to convert them into a system of coupled nonlinear ODEs. These ODEs are solved by applying the method of Quasi-Linearization, and on the other hand, structural derivatives are approximated by using the central difference technique. The effects of employing parameters on temperature and velocity are graphically studied. During the injunction, mass transfer decreases and increases in the case of suction transfer rate of heat and mass for the Reynolds number.

AI-Powered Pest Detection and Management

Muhammad Sajid Farooq — 2025-07-24

Agricultural pest infestations cause significant crop losses globally, with an estimated 20-40% of annual yield reduction attributed to pest damage. Traditional pest management methods, reliant on periodic field scouting and uniform pesticide application, are often inefficient, costly, and environmentally harmful. This paper presents an AI-powered pest detection and management system that combines deep learning-based image recognition with IoT-enabled field monitoring to enable early, accurate, and targeted pest control. The proposed system utilizes convolutional neural networks (CNNs) trained on a dataset of 25,000+ field images spanning 15 major crop pests, achieving 96.2% detection accuracy, outperforming conventional methods by 31%. Deployed on edge devices with real-time image processing capabilities, the system identifies pest hotspots and species through smartphone or drone-captured images. Integrated with wireless sensor networks, it monitors microclimatic conditions (temperature, humidity) that influence pest outbreaks, enabling predictive analytics for infestation risks.

Predicting Crop Yields Using Satellite Data and ML

Muhammad Shoukat Aslam — 2025-07-24

Agribusiness planning, economic security, and world food security lie in the proper prediction of crop yields. The research paper introduces a sophisticated machine learning model that exploits the application of satellite imagery and climatic data in order to accurately predict crop yields. This system combines the multi-spectral satellite data (Sentinel-2 2, Landsat 8 ) measuring the most important vegetation indices (NDVI, EVI ) with the weather variables (precipitation, temperature, soil moisture ) to provide accurate yield predictions several weeks before the harvest. We fit XGBoost, Random Forest, and Long Short-Term Memory (LSTM) machines and perform a combination of machine learning techniques altogether based on the hybrid approach. The model, trained with five years of data in three large corn fields in the USA (corn, wheat, soybean), has an accuracy prediction of 92.4 percent (R2 score) with regards to predictions of corn yield, 27 percent better than conventional models. Because the system offers early yield predictions (8-12 weeks before harvest) at less than 10% average relative error, profound yield-limiting parameters, including drought tension and nutrient shortages, may also be detected. The cloud-based design of the framework allows scalable deployment and thus is available to large-scale agribusiness as well as to smallholder farmers. The usage advantages of field validations include precision farming, point-to-point product market forecasting, and climate response strategy. The study finds application in the sustainable intensification of food production in the sense that it would provide information that would be used in making agricultural decisions optimally.

AI for Automated Plastic Waste Sorting

Muhammad Hassan Ghulam Muhammad — 2025-07-24

Increased plastic waste is becoming a problem that needs new ways of handling recycled waste more efficiently. The current sorting systems based on labour and near-infrared spectroscopy suffer from material variation and the scalability of operation. The given paper addresses the topic of applying artificial intelligence in automatically classifying plastic waste, suggesting a deep learning architecture that incorporates both spectral and image-based analyses. The system uses convolutional neural networks to analyze hyperspectral imagery with the goal of a higher rate of identifying the most typical polymers and combines the limitations of the traditional ones. The aim of industrial feasibility is a modular robotic interface. This work presents the value of AI-based systems in minimizing human involvement in the process of sorting waste and preconditioning sustainable recycling systems.