Multiyear Energy System Planning via Adaptive Year-Level Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Microgrid and Distributed Energy Resources (DER) project planning methods fail to effectively consider multiple-year forecasts of grid conditions, technology, fuel, electricity pricing, regulatory constraints, and climate conditions, leading to unrealistic assumptions and significant economic and environmental risks due to the inability to accurately model and optimize long-term scenarios within a reasonable time frame.
Innovation Solution
A fast multiyear projection planning method using optimization, simulation, or modeling that applies projection factors to input data to represent expected future conditions, allowing for optimal asset selection, sizing, and dispatch across an infinite number of years, utilizing machine learning and cloud-based platforms to reduce computational complexity and improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the forward-looking approach is used to model, simulate, or optimize the entire time horizon simultaneously, then future conditions are properly considered, but the time required increases non-linearly with the number of years
Solution Approach 1:
The patent segments the multiyear optimization problem into individual year-level optimization problems. Each year is solved separately using the adaptive method, where the system state is updated iteratively from one year to the next. This segmentation reduces computational complexity from exponential growth to linear scaling with the number of years, while still capturing future conditions through the adaptive updating mechanism.
2Reliability
If stochastic elements are added to hedge against uncertainty, then robustness against forecast errors improves, but the number of decisions to be made increases, further increasing runtime
Solution Approach 1:
The patent changes the approach from stochastic optimization (adding probabilistic elements and multiple scenarios) to adaptive deterministic optimization. Instead of modeling uncertainty through multiple stochastic scenarios, the adaptive method uses updated system states and forecasts for each year, achieving robustness through temporal adaptation rather than probabilistic multiplication of decision variables.
3Productivity
If simplifications are employed to combat time increase, then computational speed improves, but the reality of economic planning is distorted
Solution Approach 1:
The patent introduces dynamics into the optimization process by allowing the system state to adapt and evolve from year to year. Rather than using static simplifications that freeze parameters, the adaptive method dynamically updates forecasts, system states, and optimization results for each year, maintaining economic planning accuracy while achieving computational efficiency through sequential rather than simultaneous solving.
4Loss of time
If only a single year is solved with constant conditions assumed (SYO), then computational time is reduced, but input data is frozen for the entire project lifetime which is unrealistic
Solution Approach 1:
The patent ensures continuity of useful action by iteratively solving year-level optimization problems and updating the system state continuously from one year to the next. This continuous adaptation allows the system to capture changing future conditions over the entire project lifetime, unlike single-year optimization that freezes data, while maintaining computational efficiency through sequential solving rather than simultaneous multiyear optimization.
Data Source
AI summary
A machine learning based multiyear projection planning for energy systems is disclosed. In some embodiments, a method comprises: obtaining input data for an energy system; determining one or more projection factors based on the input data; determining, based on a machine learning model, an operation or investment associated with the energy system to achieve lower cost or improve one or more metrics of the energy system for the multiyear horizon based at least in part on the one or more projection factors and a description of technology or infrastructure of the energy system; generating a recommended operation or investment decision for the energy system based at least in part on output of the machine learning model; and storing the recommended operation or investment decision.