Advance Research in Power Electronics and Devices

Optimization and Performance Analysis of Modelled Rooftop Solar Photovoltaic Systems for Urban Energy Sustainability

2026-04-03T04:40:56+00:00

The rising demand for sustainable energy in urban areas has intensified the need for efficient rooftop solar photovoltaic (PV) systems. Urban rooftops are often constrained by limited space, shading from surrounding buildings, and variable electricity demands, which can reduce the performance of PV installations if not properly accounted for. This study presents a simulation-based approach to optimize rooftop PV systems, integrating module and inverter specifications with realistic urban load profiles to enhance energy output, system efficiency, and self-consumption. The PV system model incorporates monocrystalline silicon modules rated at 330 W with 20.1% efficiency and a temperature coefficient of -0.4%/°C, paired with 5 kW inverters featuring 97.5% efficiency and a maximum DC input voltage of 600 V. Load demand data for a typical day, including both base and variable loads, were used to simulate system performance across 24 hours. The Particle Swarm Optimisation (PSO) algorithm was applied to determine the optimal module quantity, tilt, and orientation, thereby maximising total annual energy generation while aligning production with building load patterns. Results demonstrate that the optimized configuration improves total annual energy output by X% relative to the baseline, increases system efficiency from Y% to Y1%, and raises the self-consumption ratio to A1%, reducing reliance on the central grid. The study highlights the effectiveness of simulation-based modelling combined with optimisation techniques for designing rooftop PV systems that can meet urban energy sustainability goals.

A Review on Electrical Switches Utilized in Automation with Emphasis on Working Mechanism and Industrial Usage

2026-03-02T11:48:02+00:00

Power transformers are critical assets in electrical power distribution systems, and their unexpected failure can lead to severe power outages, economic losses, and safety hazards. Conventional transformer monitoring techniques rely mainly on periodic inspections and offline testing, which are inadequate for early fault detection and preventive maintenance. This study presents a comprehensive review and implementation framework of an IoT-based smart transformer health monitoring system aimed at enabling continuous, real-time condition assessment. The proposed system focuses on monitoring key transformer parameters such as temperature and oil level using appropriate sensors interfaced with a microcontroller-based control unit. Real-time data is locally displayed and analysed using predefined threshold logic to identify abnormal operating conditions. Automatic local control actions, including relay-based cooling fan activation and audible fault alerts, are incorporated to mitigate thermal stress and prevent insulation degradation. In addition, IoT-enabled wireless communication allows real-time data transmission to a cloud platform for remote monitoring, alert generation, and historical data analysis. The observed results demonstrate reliable parameter monitoring, timely fault detection, effective temperature control, and stable cloud connectivity. The system supports predictive and condition-based maintenance strategies while maintaining low implementation cost and scalability. Overall, this work highlights the effectiveness of integrating sensing, automated control, and IoT technologies to enhance transformer reliability, operational safety, and asset management in modern power distribution and smart grid environments.

A Study into Accurate Blood Pumping in Motor-powered Artificial Hearts

2026-02-17T08:12:00+00:00

The advancement of Artificial Heart (AH) technology depends on the creation of biocompatible, energy-efficient pumping mechanisms. To mimic the physiological characteristics of natural ventricles while reducing the drawbacks of traditional axial and pulsatile systems, this study investigates a motor-driven centrifugal pump. In order to modify rotational speed (RPM) and provide biomimetic cardiac output, the suggested system uses a brushless DC motor with closed-loop control that is combined with pressure and flow sensors. The motor-driven pump achieves a mean flow rate of 5.2 L/min at 3000 RPM with a hemolysis index of 0.05 g/dL and thrombogenic potential lowered by 40% in comparison to current AH models, according to Computational Fluid Dynamics (CFD) simulations and in vitro experiments using a blood analog fluid. According to energy consumption measures, a lithium-polymer battery can run for 12 hours at a rate of 8.3 W. Under simulated rest and exercise situations, the system's adaptive feedback mechanism which modifies motor load in response to real-time hemodynamic demands was validated, guaranteeing that stroke volumes remained within physiological ranges (50–80 mL). These results highlight the potential of the motor-integrated pump to improve cardiac output while maintaining erythrocyte integrity and reducing energy overdraw, hence resolving important issues with existing AH platforms.

Intelligent State of Charge Estimation of Lithium-ion Batteries Using Machine Learning and Deep LSTM Networks: A Comparative Study

2026-03-17T07:10:27+00:00

Accurate State‑of‑Charge (SoC) estimation is a key function of the battery management system in electric vehicles, directly influencing safety, usable range, and power‑limit decisions. This study presents a comparative data-driven framework for SoC estimation of an LG 18650HG2 lithium-ion cell using three representative machine-learning architectures: ensemble regression trees, a feedforward Artificial Neural Network (ANN), and a long short-term memory (LSTM) recurrent neural network. A publicly available high-fidelity dataset comprising dynamic automotive drive-cycle profiles and multiple temperature conditions is employed, with systematic preprocessing and feature construction. All models are implemented in a MATLAB environment and evaluated using a unified protocol based on RMSE, MAE, and R² metrics, together with error histograms, residual analysis, cumulative error distributions, SoC‑band and temperature‑band robustness, and inference‑time measurements. The results establish a clear performance hierarchy: the LSTM achieves the highest accuracy with an RMSE of 1.44%, followed by the shallow ANN with 1.72%, while ensemble trees exhibit significantly larger error at 3.66% RMSE. The LSTM further delivers automotive‑grade reliability, with 96% of test predictions within ±2% SoC error and consistent performance across SoC and temperature ranges.

Controller-based Design and Real-time Implementation of a Solar Tracking Unit

2026-02-18T05:14:13+00:00

The efficiency of Photovoltaic (PV) installations can be significantly improved by ensuring that solar panels are consistently oriented to receive the maximum sunlight throughout the day. To achieve this, a single-axis automated solar tracking system has been developed, which dynamically adjusts the panel’s angle based on real-time light intensity variations. The system utilizes Light Dependent Resistor (LDR) sensors to detect the direction of the strongest illumination. A servo motor, controlled by sensor feedback, repositions the solar panel to follow the sun’s movement from sunrise to sunset. This continuous alignment enhances solar energy capture and results in higher power generation compared to traditional fixed-mount panels. To further optimize energy conversion, a Maximum Power Point Tracking (MPPT) controller is integrated into the system. It ensures efficient power transfer, minimizes energy losses, and maintains optimal charging of a connected lithium-ion battery. The entire operation is managed by an Arduino Uno microcontroller, which processes sensor data, controls motor movement, and coordinates energy optimization tasks with high responsiveness and precision. The proposed solar tracking setup offers improved energy yield, reliable battery storage, and robust performance under varying environmental conditions. Its compact, low-power design makes it ideal for portable solar charging stations, standalone renewable energy units, and smart power applications in remote or mobile settings. This system exemplifies intelligent energy harvesting for sustainable technology solutions.