Solar Irradiance Forecasting via Cloudiness Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The intermittency of solar power generation due to cloudiness and local weather conditions poses challenges for predicting solar power output, making it difficult for grid operators to maintain a balanced electricity supply and demand, as existing energy storage solutions have limitations for large-scale implementation.
Innovation Solution
A method using unsupervised learning to classify cloudiness based on historical irradiance data, updated with supervised learning using historical weather data, predicts irradiance by employing a regression model associated with the cloudiness state, allowing for accurate solar output forecasting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If solar power is integrated into the grid, then renewable energy supply increases, but intermittency and unpredictability worsen
Solution Approach 1:
The system performs preliminary classification of sky conditions into distinct cloudiness states using unsupervised learning on historical irradiance data, then applies supervised learning to associate these states with weather data. This preliminary categorization enables more accurate forecasting by preparing classification models in advance that can quickly predict irradiance patterns based on current sky conditions.
Solution Approach 2:
The patent transforms continuous irradiance data into discrete cloudiness state classifications through unsupervised learning, then uses these classified states as parameters to select appropriate regression models. This parameter transformation from continuous to discrete states improves predictability by grouping similar weather patterns together.
2Reliability
If energy storage solutions are implemented, then grid balance is maintained, but scalability and cost-effectiveness worsen
Solution Approach 1:
The patent replaces physical energy storage mechanisms with an information-processing system that uses machine learning algorithms to predict solar irradiance. By substituting mechanical storage solutions with computational forecasting models, the system maintains grid balance through better prediction rather than through physical storage capacity.
3Measurement precision
If cloudiness classification is performed, then irradiance prediction accuracy improves, but computational complexity worsens
Solution Approach 1:
The patent segments the continuous spectrum of sky conditions into distinct cloudiness states through unsupervised learning classification. By dividing complex weather patterns into manageable categories (clear, partially cloudy, overcast, etc.), the system improves prediction accuracy while making the computational problem more tractable through structured segmentation.
Solution Approach 2:
The patent introduces cloudiness state classification as an intermediary layer between raw weather data and irradiance prediction. This intermediate classification step simplifies the relationship between input data and output predictions, making the overall system more interpretable and computationally efficient.
Data Source
AI summary
Methods and systems for predicting irradiance include learning a classification model using unsupervised learning based on historical irradiance data. The classification model is updated using supervised learning based on an association between known cloudiness states and historical weather data. A cloudiness state is predicted based on forecasted weather data. An irradiance is predicted using a regression model associated with the cloudiness state.