Solar Energy Forecasting via Sky Image Pixel Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating and forecasting solar energy production inaccurately account for the location of clouds relative to solar systems, leading to discrepancies in predicted and actual radiation, which affects energy management and optimization.
Innovation Solution
A method that involves receiving images of the sky, extracting pixel values, determining separation distances to optical markers, generating a vector of parameters for pixel classification, comparing these parameters with predetermined energy production indicators, and weighting them based on distance to estimate energy production more precisely.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If cloud cover impact is considered independently of position relative to area of interest, then processing simplicity is maintained, but measurement precision deteriorates due to significant discrepancies between estimated and actual ground radiation values
Solution Approach 1:
The patent applies local quality by differentiating the impact of cloud covers based on their position in the sky relative to the area of interest. Pixels are classified into different categories (first category for cloud covers closer to the area of interest, second category for cloud covers farther away) and assigned different weights accordingly. This allows the system to account for the fact that cloud covers at different positions have different influences on ground radiation, thereby improving measurement precision without significantly complicating the processing framework.
2Measurement precision
If distance-based weighting of pixels is implemented, then energy production estimation accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent changes the parameter of pixel weighting from uniform to distance-based. Each pixel's weight is determined by its distance from the optical marker, with pixels closer to the marker (representing sky regions closer to the area of interest) receiving higher weights. This parameter change enables more accurate energy production estimation by reflecting the actual geometric relationship between sky elements and the ground area, while the computational increase is managed through efficient distance calculation and classification algorithms.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
The invention relates to a method for estimating and forecasting an energy production indicator for a solar system. The method comprises, in particular, the steps of: a) receiving (300) images of the sky taken from the ground, each image comprising an optical marker (210, 212); b) extracting values of pixels and locating of image pixels from the received images; c) determining (310) distances between the located pixels and the optical marker of the image; d) generating (320) a vector of parameters from a pixel classification, according to the extracted values and the determined distances; e) comparing (330) parameters of the generated vector with predetermined parameters, the predetermined parameters being respectively associated with energy production indicators; f) estimating (360) the energy production indicator of the solar system from energy production indicators of the predetermined parameters compared to the generated vector.