Neural Network Energy Storage Control for Peak Demand
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional utility models are inefficient in managing peak demand, leading to high costs and pollution, as they require excess capacity to meet peak electricity demands, which occur only briefly, and existing forecasting methods like regression analysis struggle to accurately predict coincident peaks due to their complexity and non-linearity.
Innovation Solution
A system utilizing a deep neural network-based control system that forecasts utility power loads by combining historical load data, weather forecasts, and real-time economic factors to identify potential coincident peaks, allowing energy storage systems to reduce energy draw from the grid during peak hours by consuming stored energy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional utility models use excess capacity to meet peak demand, then reliability of power supply is improved, but cost and pollution increase significantly
Solution Approach 1:
The system performs preliminary actions by charging energy storage systems during off-peak hours before peak demand occurs. The neural network forecasts peak events in advance, allowing the system to pre-position stored energy that will be discharged during coincident peaks, thereby meeting peak demand without requiring utility excess capacity and avoiding the associated energy waste and pollution.
Solution Approach 2:
The energy storage system acts as an intermediary between the utility grid and the electrical infrastructure. It absorbs excess energy during off-peak periods and releases it during peak periods, mediating the mismatch between utility supply capabilities and peak demand requirements. This intermediary function eliminates the need for utility peaker plants and reduces overall energy waste.
2Device complexity
If traditional forecasting methods like regression analysis are used to predict coincident peaks, then simplicity of the forecasting model is maintained, but prediction accuracy deteriorates due to complexity and non-linearity of load patterns
Solution Approach 1:
The patent replaces traditional mechanical/mathematical regression analysis with an artificial neural network system. The neural network uses parallel distributed processing and adaptive learning to capture non-linear relationships in power load data, achieving superior prediction accuracy for coincident peaks. This substitution transforms the forecasting approach from simple but inaccurate linear methods to a more complex but highly accurate adaptive system.
Data Source
AI summary
A control system for controlling an energy storage system includes a controller including a plurality of layered nodes configured to form an artificial neural network trained to generate a forecasted transmission level load and confidence value for an entire jurisdiction of a utility distribution system. The controller includes at least one memory and at least one processor configured for: identifying a potential coincident peak for the utility distribution system based on the forecasted transmission level load and confidence value generated by the artificial neural network; and upon identifying a potential coincident peak, transmitting signals to cause the electrical infrastructure to consume energy stored at the energy storage system thereby reducing the energy drawn from the utility distribution system during the identified potential coincident peak.