Solar PV-QSO: a Quantum Swarm Optimization Framework for Enhancing Photovoltaic Power Efficiency
Year: 2026
Type: conference-paper
Abstract: Artificial Intelligence (AI) integration with renewable energy systems has significantly contributed to the optimization of solar photovoltaic (PV) power. This research introduces Solar PV-QSO, an innovative QSO algorithm to optimize the efficiency of the photovoltaic cell by an intelligent MPPT. Classical Particle Swarm Optimization (PSO), Incremental Conductance (INC), and Perturb and Observe (P&O) algorithms have limitations, such as slow convergence, vulnerability to partial shading, and steady-state oscillations. Solar PV-QSO solves these problems by taking advantage of quantum-inspired methods to encode particles and maintain adaptive rotation gates to search the space. The framework follows the modality of high-resolution acquisition of the irradiance, temperature, humidity, and PV output data provided by IoT sensors, preprocessing with Z-score norms and feature selection using mutual information filtering. The QSO algorithm affects the duty cycle of a DC-DC converter dynamically to track the MPP on a real-time basis. Its performance is evaluated through simulation and real-time execution on Raspberry Pi device, and results clearly show that Solar PV-QSO produces an average tracking accuracy of 99.82, superior to classical PSO (97.43), INC (94.98), and P&O (92.87). Moreover, it brings convergence time down by more than 35% and generates 15% more energy per day. The suggested framework is lightweight with < 100 MB requirements in terms of memory and < 21% utilization of the CPU. Such findings point to the possibility of scalable and cost-effective implementation in smart micro grids, off-grid solar and adaptive renewable infrastructure, particularly facing fastchanging environmental conditions.
Institutions: Gnanamani College of Technology, Amrita Vishwa Vidyapeetham
Language: English
Field: Energy
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204835228