一种基于无人机的低空数据智能采集分析方法
By optimizing the flight control of UAVs through multimodal data reconstruction and joint state equations, the problems of failed data recovery and excessive energy consumption caused by UAVs blindly approaching in complex low-altitude environments were solved. This achieved precise data recovery and minimized energy consumption, improving the reliability of data acquisition and the fitting accuracy of the model.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Filing Date
- 2026-04-24
- Publication Date
- 2026-07-17
AI Technical Summary
In complex low-altitude environments, drones lack rigorous mathematical quantification constraints on atmospheric optical attenuation characteristics and gaps in underlying visual information. This leads to blind approach strategies resulting in failed re-extraction or excessive approach causing battery overdraft. It is impossible to accurately define a safe and effective approach distance.
Multimodal data is acquired through airborne sensors. The latent space features of the first modal data are reconstructed using the second modal data. A pixel-level uncertainty variance matrix is output. A joint state equation set is established by combining information inverse and extinction coefficient. A Hamiltonian function that minimizes flight energy consumption is constructed. The optimal thrust control law and the approach stop boundary are output. The UAV is driven to perform precise re-sampling and trigger the airborne sensors to acquire true data. Finally, the true data is used for fine-tuning of model parameters.
It achieves precise re-extraction and minimizes energy consumption in degraded environments, improves the re-extraction failure and power depletion problems caused by the traditional blind approach strategy, and enhances the model's fitting accuracy for specific degraded weather conditions and the reliability of data acquisition.
Smart Images
Figure CN122411528A_ABST