PV Irradiance Prediction Correction via Reference Curves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current photovoltaic power systems face instability and inefficiency due to inaccurate short-term cloud predictions, leading to false alarms and reduced solar energy output, as existing geostationary satellite-based forecast methods lack spatial and temporal resolution for very short-term forecasts, and computer vision techniques struggle with sky image segmentation, especially near the sun where glare and optical artifacts are present.
Innovation Solution
A system and method that uses direct visual observations and analysis to estimate short-term solar irradiance by generating an observed irradiance curve, a reference curve for clear sky conditions, and correcting predictions through probabilistic analysis of time-segmented intervals, reducing false alarms and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If geostationary satellite-based forecast approaches are used, then coverage area is improved, but spatial and temporal resolution deteriorates
Solution Approach 1:
The sky is divided into multiple regions (e.g., regions of interest around the sun, other sky regions) that are analyzed separately. Image segmentation techniques partition the sky image into cloud pixels and clear sky pixels, allowing detailed local analysis while maintaining overall coverage.
Solution Approach 2:
The system transitions from satellite-based 2D imaging to ground-based 3D hemispherical sky imaging. By capturing sky radiation in multiple directions and analyzing cloud position, size, and movement in three-dimensional space, the system achieves higher spatial and temporal resolution for short-term forecasts.
2Extent of automation
If computer vision techniques are used for sky image segmentation, then automation is improved, but measurement precision deteriorates due to glare and optical artifacts
Solution Approach 1:
Different classification approaches are applied to different regions of the sky. Regions near the sun with glare and optical artifacts use specialized handling, while other regions use standard classification. The system adjusts classification thresholds and methods based on local sky conditions.
Solution Approach 2:
The system introduces intermediate processing steps between raw image capture and final cloud classification. This includes glare detection and masking, optical artifact identification and removal, and multiple-stage classification processes that refine cloud detection accuracy.
3Stability of the object's composition
If short term cloud forecast is implemented, then power generation stability is improved, but false alarms increase due to inaccurate predictions
Solution Approach 1:
The system continuously monitors actual sky conditions and compares them with forecast predictions. By analyzing the difference between predicted and actual cloud positions and movements, the system refines its forecasting algorithms and reduces false alarms over time.
Solution Approach 2:
The system performs preliminary analysis of sky images and cloud patterns before final prediction. By identifying potential cloud movements and assessing their likelihood of reaching the sun in the near future, the system can distinguish between actual threats to power generation and false alarm conditions.
Data Source
AI summary
A system and method for correcting short term irradiance prediction for control of a photovoltaic power station includes an irradiance data manager that generates an observed irradiance curve based on irradiance measurements received from an irradiation sensor, and a reference curve generation engine that generates a reference curve representative of irradiance values for a clear sky day. An irradiance prediction engine generates an irradiance prediction curve based on image segmentation and motion filtering of cloud pixels in sky images. A prediction correction engine corrects the irradiance prediction curve within a short term future time interval based on probabilistic analysis of time segmented intervals of the observed irradiance curve, the reference curve, and the irradiance prediction curve.