MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641109622 A) filed by St. Josephs College Of Engineering Kanchipuram, Tamil Nadu; Sumashkrishna M; Vijay Vigneswar K R; and Mr. Prasad H on September 12, 2026, for Analysis Of Electrical Parameter Fault Detection For Maintenance In Solar Pv Systems Using Machine Learning.

Inventors include St. Josephs College Of Engineering Kanchipuram, Tamil Nadu; Sumashkrishna M; Vijay Vigneswar K R; and Mr. Prasad H.

The application for the patent was published on September 18, 2026, under issue no. 38/2026.

Abstract: ABSTRACT Solar photovoltaic systems are prone to gradual performance degradation due to environmental and operational factors such as dust accumulation, temperature variations, shading, aging, and inverter faults. These issues often lead to silent efficiency losses that remain undetected until significant energy reduction occurs. Traditional maintenance methods, which rely on periodic inspections or reactive fault handling, are inefficient and increase operational costs. This project proposes a Machine Learning based predictive maintenance system for solar panels that enables continuous monitoring and early fault detection. The system utilizes key parameters such as solar irradiance, temperature, voltage, current, and weather data to model the normal behavior of solar panel output. A regression-based approach is employed to predict the expected power generation, and deviations between predicted and actual output are analyzed to identify anomalies. A statistical thresholding mechanism is used to detect abnormal performance, allowing the system to generate timely alerts and indicate potential causes such as soiling, degradation, or system faults. The proposed approach helps in minimizing downtime, reducing maintenance costs, and improving overall energy efficiency. The project demonstrates a scalable and cost-effective solution for intelligent solar monitoring, with potential extensions including IoT integration, advanced time series models, and deployment in large-scale solar energy systems.

Disclaimer: Curated by HT Syndication.