Solar Panel Infrared Defect Detection for Microcracks and Hotspots
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting defects in solar panels, such as manual inspections and imaging techniques, are inefficient, labor-intensive, and lack accuracy, particularly in identifying microcracks and early-stage hotspots, which can worsen over time and reduce the performance of solar power plants.
Innovation Solution
A method and system using a You Only Look Once Series—Photovoltaic (YOLOS-PV) convolutional neural network with a pretrained vision transformer and a transformer block encoder to process infrared images from solar panels, applying position embedding, generating detection tokens, and classifying defects with multi-layer perceptron heads, utilizing a weighted loss combination of GIoU and L1 regression loss for precise defect detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual inspection is used to detect defects in solar panels, then reliable assessment can be provided, but the process becomes time-consuming and labor-intensive
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated imaging system that captures photographs of solar panels and uses machine learning algorithms to detect defects. The system substitutes human visual inspection with computational image analysis, achieving both high reliability and efficiency simultaneously.
2Reliability
If conventional thermal and electrical modeling is used to assess solar panel performance, then temperature variations and power output can be analyzed, but structural defects like microcracks cannot be identified
Solution Approach 1:
The patent creates a multi-functional defect detection system that can identify multiple types of defects simultaneously - both functional defects (hotspots, performance issues) and structural defects (cracks, broken glass) - using a single integrated imaging and machine learning platform, eliminating the need for separate specialized equipment.
3Measurement precision
If advanced imaging techniques such as Electroluminescence or Infrared Thermography are used for defect detection, then enhanced visibility of defects can be achieved, but the techniques require specialized equipment and controlled conditions
Solution Approach 1:
The patent uses standard digital cameras or smartphones to capture images of solar panels, replacing expensive specialized imaging equipment like Electroluminescence or Infrared Thermography systems. The approach uses readily available, inexpensive devices combined with advanced software algorithms to achieve comparable or superior detection accuracy.
4Measurement precision
If signal processing techniques such as wavelet transform are applied to enhance defect detection, then computational analysis can be improved, but the techniques involve high computational costs and lack generalization
Solution Approach 1:
The patent applies pre-trained machine learning models that have already learned defect patterns during an offline training phase. During actual inspection, the pre-trained model quickly processes images without requiring intensive real-time computation, achieving both high precision and low energy consumption during deployment.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The YOLOS-PV convolutional neural network enhances defect detection accuracy and automation, effectively identifying various defects in solar panels, reducing false positives, and ensuring timely maintenance for improved efficiency and longevity.
Implementation Method 1
collecting, by a thermal camera on at least one imaging drone navigating over at least one mounted and operational solar panel, an image dataset of infrared images
Data Source
AI summary
A method and a system for detecting defects in solar panels. The method includes collecting an image dataset of infrared images of the at least one mounted and operational solar panel; applying the image dataset to a YOLOS-PV convolutional neural network; converting the image dataset to a set of patches; applying position embedding to each patch and generating a set of position embedded patches; applying the set of position embedded patches to a transformer block encoder; generating a set of detection tokens; applying the set of detection tokens to an array of multi-layer perceptron (MLP) heads; classifying defects in the position embedded patches; setting, by the MLP heads, a plurality of Bpred bounding boxes for each of the set of position embedded patches; and identifying a solar panel in need of repair based on the defects in the set of position embedded patches and the plurality of Bpred bounding boxes.