CASE C39 · IN · 2021
印度泰米尔纳德邦 55 MW 太阳能电站：热成像无人机勘察在约 233,000 块组件中标记出 2,481 块
一个研究团队用热成像无人机飞越南印度一座 55 MW 光伏电站，在约 233,000 块组件中标记出 2,481 块疑似异常组件。正射影像图作为数字孪生用于定位。未测量辐照度，文中所称的 97% 准确率也缺乏数据支撑。
| 业主 | Not named (site survey supported by Watsun Infrabuild Pvt. Ltd.) |
| 服务商 | NIT Karnataka (Surathkal) with the National Observatory of Athens (research team) |
| 地点 | Moongiltholuvu, Tamil Nadu, India |
| 机型 | Not specified (multirotor with thermal camera, flown at 130–140 ft) |
| 传感器 | thermal infrared, RGB for orthomosaic |
引入前
在如此规模下逐块从地面检查组件并不现实；高气溶胶负荷、灰尘和季风天气加速了积灰和劣化。
引入无人机后
2021年2月执行了十六条航线；用开源 WebODM 构建的正射影像图作为数字孪生，热点在 QGIS 中标记并与地面测量进行比对。
成果
| 指标 | 数值 | 依据 |
|---|---|---|
| panels mapped | 232,934 panels identified across six blocks (5 × 10 MW, 1 × 5 MW) | 报告值 |
| defects flagged | 2,481 panels (about 1.1%) indicated as defective (hotspots from manufacturing faults, ageing, shading, soiling, hail) | 报告值 |
| accuracy | "up to 97%" stated in conclusions; underlying validation numbers not shown | 企业宣称 |
注意事项与局限
- Single-site academic study; the plant owner is not named and drone/camera models are not given.
- The abstract says defects were indicated automatically, but the methods describe manual marking; the 97% accuracy figure is not traceable in the text.
- Irradiance was not measured during the flights, although thermography guidance (about 500–700 W/m²) depends on it. The authors cite a lack of thermographic know-how as a main limitation.
如果您还想了解无人机自身全生命周期的排放，见物流指南的论文卡片Lifecycle modeling and assessment of unmanned aerial vehicles (drones) CO2e emissions。
来源
- Solar Photovoltaic Hotspot Inspection Using Unmanned Aerial Vehicle Thermal Images at a Solar Field in South India (Remote Sensing 15(7):1914)MDPI Remote Sensing 已读全文
来源已核对: 2026-10-22