一种基于组件轮廓视觉提取的清扫机器人纠偏方法
By performing specific processing on the RGB images of photovoltaic modules and calculating the visual yaw factor, combined with decoupled control using attitude modulation factors, the problems of low recognition accuracy and control oscillation of the cleaning robot in the photovoltaic power station environment were solved, achieving high-precision path correction and stable walking.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SKYSYS INTELLIGENT TECH SUZHOU CO LTD
- Filing Date
- 2026-03-10
- Publication Date
- 2026-07-17
AI Technical Summary
Existing cleaning robots suffer from low recognition accuracy due to environmental noise in photovoltaic power plant environments, control oscillations caused by sensor jumps, and unstable navigation path correction, making it difficult to achieve efficient, safe, and intelligent operation and maintenance.
By acquiring RGB images of photovoltaic modules, performing specific channel weighting processing and inverse logarithmic gain adaptive compensation, and combining the prior geometric constraints of robot movement to select straight lines, calculating visual yaw factor and lateral displacement trend degree, and introducing attitude modulation factor for decoupling control, attitude correction takes precedence over position correction.
It improves the saliency and recognition accuracy of photovoltaic module texture features, solves the problem of sensor reading jumps, enhances the robot's walking stability and path correction accuracy in complex environments, and eliminates control oscillation and serpentine walking problems.
Smart Images
Figure CN121797704B_ABST