An Intelligent Operation and Maintenance System for Photovoltaic Station

Abstract

The rapid expansion of photovoltaic (PV) deployment poses new challenges for large-scale and distributed maintenance, particularly in fishery-PV complementary plants where panels are deployed over water surfaces. This paper presents the design and implementation of an intelligent operation and maintenance (O & M) system that integrates a 3D holographic digital twin cloud platform with UAV-assisted inspection and localized cleaning. The proposed system supports multi-source data acquisition, including UAV imagery, infrared sensing, and DustIQ-based soiling monitoring, and provides real-time visualization of the PV plant through 1:1 3D reconstruction. UAVs are employed for both autonomous inspections, covering defects such as soiling, bird droppings, bypass diode faults, and panel disconnections and targeted cleaning in small water-covered areas. Field trials were conducted at Riyue and Chebu PV plants, with small-scale UAV cleaning validation in Chebu fish ponds. Results demonstrated that the system achieves efficient task scheduling, fault detection, and localized cleaning, thereby improving O & M efficiency, reducing costs, and enabling digitalized and intelligent management for large-scale PV stations.