Neural Network Ensemble for Renewable Energy Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Renewable power generation plants face challenges in predicting and categorizing energy output due to environmental variations, leading to fluctuating energy infeeds that are not effectively utilized in energy transmission networks.
Innovation Solution
A computer-assisted method using an ensemble of neural networks, each with different structures and training, to forecast future energy outputs and calculate confidence levels for energy availability, allowing for appropriate allocation and pricing of energy amounts based on likelihood, including the use of weather data and geographic parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If weather forecasts are used to determine predictions for future energy production, then energy planning can be improved, but the energy produced is not categorized for different purposes leading to inefficient utilization
Solution Approach 1:
The patent segments the forecasted energy production into different categories based on confidence levels (e.g., high confidence, medium confidence, low confidence). This segmentation allows different energy amounts to be allocated to different uses - high confidence energy for guaranteed delivery, medium confidence for flexible demand, and low confidence for reserve or curtailment scenarios.
Solution Approach 2:
The patent changes the parameter of energy categorization from a single aggregate prediction to multiple categorized predictions with associated confidence levels. This transformation enables the energy management system to make differentiated decisions based on the reliability of predictions, improving overall energy utilization efficiency.
2Device complexity
If a single neural network is used for energy forecasting, then the model is simple to implement, but it cannot provide confidence levels for different energy amounts
Solution Approach 1:
The patent combines multiple neural networks into an ensemble system where each network provides a prediction. The results are then aggregated and analyzed to determine confidence levels for different energy amounts. This merging approach preserves the relative simplicity of individual neural networks while gaining the additional information about prediction confidence.
Solution Approach 2:
The patent introduces an intermediary processing step that takes the outputs from multiple neural networks and transforms them into categorized energy amounts with confidence levels. This intermediary layer synthesizes the predictions and assigns confidence levels without requiring complex changes to the underlying neural network structures.
3Ease of operation
If energy amounts are not categorized by likelihood of availability, then the system is simple to operate, but energy cannot be effectively allocated for different purposes
Solution Approach 1:
The patent performs preliminary categorization of energy amounts by confidence level before the actual energy allocation decision is made. This preliminary action organizes the energy predictions into ready-to-use categories that can be directly matched with different demand types, simplifying the subsequent allocation process while enabling versatile energy management.
Data Source
AI summary
A method for computer-assisted determination of usage of electrical energy produced by a power generation plant such as a renewable power generation plant is provided. The method uses a plurality of neural networks having a different structure or being learned differently for calculating future energy amounts produced by a power generation plant. To do so, the energy outputs of the power generation plant forecasted by the plurality of the neural networks are used to build histograms. Based on the histograms, energy amounts for different confidence levels describing the likelihood of the availability of the energy amount are determined, and different uses are assigned to different energy amounts. Energy amounts having a higher likelihood of availability in the future are sold at higher prices than other energy amounts.