This is the authors' abstract. We don't add key points for this paper.
This review provides a systematic examination of recent advances in solar photovoltaic tracking, with a critical focus on the system shift toward AI-IoT convergence as a transformative strategy for energy optimization. Through a quantitative synthesis of studies published between 2015 and 2025, we report that single-axis trackers achieve energy gains of 15%–35% over fixed-tilt installations, while dual-axis systems attain energy improvements of 30%–45%. However, net gains are mediated by latitude, with equatorial regions showing marginal economic returns despite technical superiority (e.g., 35.5% power increase in Malaysia versus 19.62% in Ecuador). The review introduces a novel analytical framework that distinguishes AI applications for tracking control (mechanical orientation) from those for broader PV optimization (MPPT, fault diagnosis, energy management), a distinction absent in prior literature. We systematically evaluate algorithmic performance across both domains. Reinforcement learning methods achieve tracking accuracy exceeding 95% with sub-second decision latencies, where deep learning architectures attain MPPT efficiencies of up to 99.89% and fault detection accuracy of 98.15%. A key contribution is the identification of the “robustness paradox”, the fundamental trade-off between rapid irradiance tracking and partial shading accuracy, and the demonstration that hybrid algorithms (e.g., the Team Game Algorithm and the Hybrid Artificial Fish Swarm Algorithm) uniquely bridge this divide, achieving 99.54% efficiency with a convergence time of 0.92 seconds. We further synthesize recent advances in generative AI for synthetic data augmentation, IoT-enabled edge-cloud architectures achieving 77% sensor reduction while maintaining 98.15% diagnostic accuracy, and explainable AI frameworks that balance transparency with performance (95.61% accuracy, 88% computational reduction). This review uniquely integrates technical performance metrics with techno-economic contextualization, providing researchers and practitioners with evidence-based guidelines for algorithm selection contingent on deployment scale, climatic regime, and resource constraints. It advances the field beyond descriptive reporting toward actionable design frameworks for next-generation resilient PV systems.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Selim et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: