AI-Guided Solar Panel Alignment Under Glare and Rotation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The installation of solar panels on photovoltaic arrays faces challenges such as ensuring all panels are coplanar and leveled on rotatable structures, and conventional methods are inefficient and costly, especially in tropical environments with glare and illumination issues.
Innovation Solution
A solar panel handling system that combines machine learning techniques with an end-of-arm assembly tool equipped with suction cups and a linear guide assembly, using force torque transducers and machine learning algorithms to detect panel centers and corners, navigate, and align panels accurately on installation structures like torque tubes, while avoiding mechanical structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional computer vision techniques are used for solar panel detection, then object detection can be performed in ideal environments, but detection accuracy deteriorates under glare, over-exposure, or under-exposure conditions
Solution Approach 1:
The patent replaces conventional computer vision techniques with a deep learning-based neural network system that processes images captured by cameras mounted on the robotic arm. This substitution enables the system to accurately detect solar panel positions, corners, and edges even under challenging lighting conditions including glare and over-exposure, thereby resolving the contradiction between detection accuracy and environmental harmful factors
Solution Approach 2:
The system changes the parameters of image processing by using a trained neural network model that has been specifically trained to recognize solar panel features under various lighting conditions. The neural network adjusts its parameter thresholds and feature extraction methods dynamically, allowing accurate detection despite glare and illumination variations that would cause conventional vision systems to fail
2Productivity
If manual installation methods are used for solar panels, then installation can be performed with simple equipment, but installation efficiency and precision deteriorate
Solution Approach 1:
The robotic system performs self-positioning and self-alignment by using its own mounted cameras to detect the positions of solar panels and installation structures. The system automatically calculates its own position relative to the torque tube and previously installed panels, then autonomously adjusts its end effector orientation and position. This self-service capability enables high installation efficiency without requiring complex external positioning infrastructure
Solution Approach 2:
The robotic arm system integrates multiple functions into a single platform: it can navigate to installation positions, detect solar panel and structure positions using onboard cameras, calculate alignment parameters, position the end effector, and perform the actual panel installation. This multi-functionality achieves high productivity while managing device complexity through integration rather than separate specialized systems
3Power
If solar panels are installed on rotatable structures to track the sun, then energy generation is improved, but ensuring all panels are coplanar and leveled becomes more difficult
Solution Approach 1:
The system uses real-time feedback from cameras mounted on the robotic arm to detect the positions of solar panels and the rotatable structure (torque tube). The neural network processes these images to determine panel orientations and the structure's rotation angle. Based on this feedback, the system calculates the precise orientation needed for the end effector to install the next panel coplanar with previously installed panels, even as the structure rotates. This closed-loop feedback control maintains manufacturing precision despite the dynamic rotatable platform
Solution Approach 2:
The system is designed to dynamically adapt to the rotating torque tube structure. Rather than requiring the structure to be static during installation, the robotic arm can accommodate different rotation angles by adjusting its own position and orientation accordingly. The neural network continuously processes visual data to track the structure's rotation and calculates real-time adjustments needed to maintain panel coplanarity, enabling the system to work effectively with dynamic rotatable installations that maximize energy generation
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 system enhances the efficiency and reliability of solar panel installation by accurately positioning and aligning panels, even in challenging environments, reducing installation costs and improving robotic handling and navigation.
Implementation Method 1
an end of arm assembly tool with which to engage a solar panel
Implementation Method 2
a linear guide assembly including a linearly moveable clamping tool with an engagement member configured to engage a clamp assembly slidably coupled to an installation structure
Data Source
AI summary
A system and method for installing solar panels are provided. The method obtains an image of a solar panel during an in-progress solar installation and estimates features of the solar panel based on a first image using distance simulation, geometric correction, and/or angular adjustment. The method also generates control signals, based on the estimated features, for operating a robotic controller for picking the solar panel. The method also obtains a second image of the solar panel when the solar panel is in a perspective view and detects placement of the solar panel based on the image by determining if the solar panel is co-planar with and at a predetermined offset from a fixed solar panel. Based on the detected placement, control signals are generated for operating a second robotic controller for aligning the solar panel with the fixed solar panel.